See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/347195122

The Evolution of Electricity Markets in Europe

Book · November 2020

DOI: 10.4337/9781789905472

CITATIONS READS

53 1,745

1 author:

Leonardo Meeus
European University Institute
172 PUBLICATIONS 2,714 CITATIONS

SEE PROFILE

All content following this page was uploaded by Leonardo Meeus on 11 April 2021.

The user has requested enhancement of the downloaded file.

The Evolution of Electricity Markets in Europe
THE LOYOLA DE PALACIO SERIES ON EUROPEAN ENERGY POLICY
Series Editor: Jean-Michel Glachant, Holder of the Loyola de Palacio Chair on EU Energy
Policy and Director of the Florence School of Regulation, European University Institute, Italy,
and Professor of Economics, Université Paris-Sud, France
The Loyola de Palacio Series on European Energy Policy honours Loyola de Palacio
(1950–2006), former Vice-President of the European Commission and EU Commissioner for
Energy and Transport (1999–2004), a pioneer in the creation of an EU Energy Policy.
This series aims to promote energy policy research, develop academic knowledge and
nurture the ‘market for ideas’ in the field of energy policy making. It will offer informed and
up-to-date analysis on key European energy policy issues (from market building to security
of supply; from climate change to a low carbon economy and society). It will engage in
a fruitful dialogue between academics (including economists, lawyers, engineers and political
scientists), practitioners and decision-makers. The series will complement the large range of
activities performed at the Loyola de Palacio Chair currently held by Professor Jean-Michel
Glachant (Robert Schuman Center for Advanced Studies at European University Institute in
Florence, Italy, and University Paris-Sud 11).
Titles in the series include:
Security of Energy Supply in Europe
Natural Gas, Nuclear and Hydrogen
Edited by François Lévêque, Jean-Michel Glachant, Julián Barquín, Christian von
Hirschhausen, Franziska Holz and William J. Nuttall
Competition, Contracts and Electricity Markets
A New Perspective
Edited by Jean-Michel Glachant, Dominique Finon and Adrien de Hauteclocque
The Economics of Electricity Markets
Theory and Policy
Edited by Pippo Ranci and Guido Cervigni
Building Competitive Gas Markets in the EU
Regulation, Supply and Demand
Jean-Michel Glachant, Michelle Hallack and Miguel Vazquez
Market Building through Antitrust
Long-term Contract Regulation in EU Electricity Markets
Adrien de Hauteclocque
Electricity Network Regulation in the EU
The Challenges Ahead for Transmission and Distribution
Edited by Leonardo Meeus and Jean-Michel Glachant
The Evolution of Electricity Markets in Europe
Leonardo Meeus
The Evolution of Electricity
Markets in Europe
Leonardo Meeus
Florence School of Regulation, European University Institute, Italy and
Vlerick Business School, Belgium

LOYOLA DE PALACIO SERIES ON EUROPEAN ENERGY POLICY

Cheltenham, UK • Northampton, MA, USA

© Leonardo Meeus 2020

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or
transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or
otherwise without the prior permission of the publisher.

Published by
Edward Elgar Publishing Limited
The Lypiatts
15 Lansdown Road
Cheltenham
Glos GL50 2JA
UK

Edward Elgar Publishing, Inc.

William Pratt House
9 Dewey Court
Northampton
Massachusetts 01060
USA

A catalogue record for this book

is available from the British Library

Library of Congress Control Number: 2020944718

This book is available electronically in the

Economics subject collection
http://dx.doi.org/10.4337/9781789905472

ISBN 978 1 78990 546 5 (cased)

ISBN 978 1 78990 547 2 (eBook)
Contents

List of figures viii

List of tables ix
List of boxes x
List of authors xi
Acknowledgements xiii
List of abbreviations xv
Introduction xx
Structure of this book xxiii

PART I HOW TO TRADE AND TRANSPORT ELECTRICITY

ACROSS NATIONAL BORDERS?

1 Why did we start with electricity markets in Europe? 2

Leonardo Meeus with Valerie Reif
1.1 What was the political process that led to electricity markets in Europe? 2
1.2 What were the technical drivers for creating a European power system? 10
1.3 What do we know about the benefits of integrating electricity markets
in Europe? 11
1.4 Conclusion 13
1A.1 Annex: Regulatory Guide 16

2 Who gets the rights to trade across borders? 25

Leonardo Meeus with Tim Schittekatte
2.1 How to deal with historical privileges? 25
2.2 How to implement market-based allocation of transmission rights? 26
2.3 How to implement market coupling in the day-ahead timeframe? 28
2.4 What about the timeframes before day-ahead? 31
2.5 What about the timeframes after day-ahead? 33
2.6 Conclusion 34
2A.1 Annex: The evolution of the role of power exchanges in Europe 38
2A.2 Annex: A simple numerical example of market coupling 40
2A.3 Annex: Regulatory guide 42

v
vi Evolution of electricity markets in Europe

3 How to calculate border trade constraints? 48

Leonardo Meeus with Tim Schittekatte
3.1 Why do we focus so much on constraints? 48
3.2 Why do we calculate trade constraints on virtual borders? 50
3.3 Who is best placed to do virtual border calculations? 51
3.4 How to organize the data exchange in support of the calculations? 52
3.5 Why are there still open issues? 55
3.6 Conclusion 57
3A.1 Annex: Flow-based market coupling (FBMC) 60
3A.2 Annex: Transit flows and loop flows 61
3A.3 Annex: Regulatory guide 62

4 Who pays for the network when trade is international? 68

Leonardo Meeus with Tim Schittekatte
4.1 Who pays for the network? 68
4.2 Why did national network tariffs start to be harmonized? 69
4.3 Why was there a move away from transit charges? 71
4.4 How to share network investment costs between countries? 72
4.5 Conclusion 74
4A.1 Annex: Congestion rent under market coupling: a numerical example 78
4A.2 Annex: Regulatory guide 79

PART II HOW TO COMBINE ELECTRICITY TRADE WITH SYSTEM

SECURITY TO KEEP THE LIGHTS ON?

5 Who is responsible for balancing the system? 84

Leonardo Meeus with Tim Schittekatte and Valerie Reif
5.1 How to share responsibility between system operators? 84
5.2 How to incentivize market parties to be balanced? 88
5.3 How to ensure that reserves are available? 89
5.4 How to integrate balancing markets across borders? 92
5.5 Conclusion 95
5A.1 Annex: Regulatory guide 100
 Contents vii

6 How to organize system operation and connection requirements? 111

Leonardo Meeus with Valerie Reif
6.1 Why pay attention to detailed technicalities? 111
6.2 How to organize regional system operation? 113
6.3 How to introduce minimum technical standards? 115
6.4 How to proceed with the implementation of technical standards? 121
6.5 Conclusion 123
6A.1 Annex: Types of grid users in the RfG and DC NC 125
6A.2 Annex: Regulatory guide 127

7 How to ensure adequate investment in power plants? 135

Leonardo Meeus with Athir Nouicer
7.1 Why did some countries introduce a capacity mechanism? 135
7.2 What is the best capacity mechanism? 137
7.3 How to limit the (ab)use of capacity mechanisms? 139
7.4 Conclusion 141
7A.1 Annex: Taxonomy of capacity mechanisms 145
7A.2 Annex: Regulatory guide 146

PART III HOW TO PUT THE CITIZEN AT THE CENTRE OF THE

ENERGY TRANSITION?

8 How to put the citizen at the centre of the energy transition? 153
Leonardo Meeus with Athir Nouicer
8.1 Why did we start this new paradigm? 153
8.2 How is this change of paradigm being implemented? 154
8.3 Conclusion 156
8A.1 Annex: Regulatory guide 158

9 Conclusion 164
Leonardo Meeus

Index 165
Figures

0.1 The sequence of electricity markets in Europe xx

1.1 The main steps in the evolution of European electricity markets 3

1.2 The development of TSOs at the national and European levels and
a selection of their tasks 6

1.3 The development of regulatory authorities at the national and European

levels and a selection of their tasks 7

1.4 A basic representation of the power system using a tandem bicycle 10

2.1 Implementation of different allocation methods for cross-border

transmission rights in Europe in 2004 26

2.2 Explicit cross-zonal allocation: hourly price difference between

Germany and the Netherlands (x-axis) versus the hourly sum of
nominated net flows from Germany to the Netherlands in 2004 (y-axis) 27

2A.1 European power exchange mergers and acquisitions (non-exhaustive) 39

3.1 Bidding zones in Europe as of November 2019 51

3.2 Left: the five regional security coordinators (RSCs) as of 1 January

2019; right: the ten capacity calculation regions (CCRs) 53

3.3 The common grid model (CGM) merging process 54

3A.1 Scheduled flows (black), transit flows (white, left) and loop flows
(white, right); the different rectangular areas represent bidding zones 61

4.1 The Norway–Sweden case (left) and the Italy–Greece case (right) 74

6.1 Capacities lost at the frequencies reached in the three areas during the
split of the Continental European system on 4 November 2006 112

6.2 Evolution of RSCIs into RSCs and then RCCs and a selection of their tasks 114

6.3 Timelines from publication to entry into application of RfG NC and DC NC 117

6A.1 Illustration of the terms ‘demand facility’, ‘distribution facility’ and

‘distribution system’ as used in the DC NC 126

viii
Tables

1A.1 Regulatory guide 16

2.1 Different design choices for the implementation of market coupling 29

2A.1 Regulatory guide 42

3A.1 Regulatory guide 62

4.1 Maximum annual average transmission charges paid by producers

according to Regulation (EU) No 838/2010 70

4A.1 A numerical example of congestion rent calculation 78

4A.2 Regulatory guide 79

5.1 Terminology for reserve products 85

5.2 Market shares of the largest BSP for FRR between 2003 and 2005 93

5A.1 Regulatory guide 100

6.1 Limits for thresholds for different types of power-generating modules 117

6.2 Frequency ranges and duration of connection requirements by

synchronous area for PGMs of all types and transmission-connected
demand facilities, transmission-connected distribution facilities,
distribution systems and demand units providing demand response services 119

6.3 Compliance of different types of power-generating modules with

a selection of technical requirements as described in the RfG NC 120

6A.1 Regulatory guide 127

7A.1 Taxonomy of capacity mechanisms 145

7A.2 Regulatory guide 146

8A.1 Regulatory guide 158

ix
Boxes

1.1 Nicknaming EU legislation 4

1.2 The EU Clean Energy Package 5

1.3 The first generation of EU electricity network codes and guidelines 8

1.4 Implications of the EU Clean Energy Package for EU network codes

and guidelines 9

3.1 The Italian blackout of 28 September 2003 49

4.1 Two examples of innovative CBCA practices 73

5.1 Primary frequency control in UCTE and frequency-controlled reserves

in NORDEL 86

6.1 RfG NC and DC NC timelines and national implementation processes 117

7.1 Rationing or load-shedding plans and controversies 136

7.2 The Tempus case 138

x
List of authors

Leonardo Meeus is Professor of Strategy and Corporate Affairs at Vlerick Business School.
He is the Director of the Vlerick Energy Centre. He is also a Professor in the Florence School
of Regulation at the European University Institute. In Florence, he is Course Director of online
courses on the latest regulatory trends in the energy sector, such as the EU Clean Energy
Package and the EU electricity network codes. He is a member of the International Association
for Energy Economics. Leonardo graduated from KU Leuven as a business engineer. He later
obtained a PhD in electrical engineering from the same university. He works as an expert for
EU institutions, regulatory agencies and companies on research contracts and in advisory
roles.

CHAPTER CO-AUTHORS

Athir Nouicer has been working as a research associate at the Florence School of Regulation
(FSR) since 2017, where he is part of the electricity regulation research team. His main
research interests are the EU Clean Energy Package and the use of flexibility in distribution
grids. He is currently also a PhD researcher at KU Leuven. Before joining FSR, Athir worked
at the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) as an energy expert
for the German–Tunisian Energy Partnership. He graduated as a mechanical engineer from
the National Engineering School of Tunis and holds two master’s degrees under the EMIN
programme: in the Digital Economy and Network Industries from Université Paris Sud XI and
in the Electric Power Industry from Universidad Pontificia Comillas.
Valerie Reif has been working as a research associate at the Florence School of Regulation
(FSR) since 2018, where she is part of the electricity regulation research team. Her main
research interests are currently the EU electricity network codes and the management and
exchange of electricity network, market and consumer data. Before joining FSR in 2018, she
worked at the technology platform Smart Grids Austria, and gathered valuable experience at
Austrian Power Grid and at the EU Representation Office of Oesterreichs Energie. Valerie
holds both a BSc and an MSc degree in Renewable Energy Engineering from the University
of Applied Sciences Technikum Wien in Austria and a BA in European Studies from the
University of Passau in Germany.
Tim Schittekatte has been a research associate at the Florence School of Regulation since
2016, where he is part of the electricity regulation research team. His main research interests
are currently EU electricity network codes, flexibility markets and distribution network tariff
design. He is also affiliated with the Vlerick Energy Centre in Brussels. Before joining FSR,
he was a visiting researcher at the Grid Integration Group at the Lawrence Berkeley National

xi
xii Evolution of electricity markets in Europe

Lab and a junior economist at Microeconomix in Paris. He graduated as an engineer from

Ghent University, Belgium and completed an international master’s in economics (the EMIN
programme). He holds a PhD in energy economics from Université Paris-Sud XI.
Acknowledgements
Like all scientific work, this book stands on the shoulders of giants. Some have been mentors
and colleagues at KU Leuven, the Florence School of Regulation or Vlerick Business School.
Others have been partners in executive education, research projects or studies for EU insti-
tutions, regulatory authorities and companies. Equally important are those who need their
work to be confirmed or challenged by academic research. We share a passion for electricity
markets and for the European integration process.
You will find the following electricity market giants in the reference lists and the endnotes,
and I also want to warmly thank them here: Carlos Batlle, Ronnie Belmans, Anna Creti,
Erik Delarue, Adrien de Hauteclocque, William D’haeseleer, Daniel Dobbeni, Christophe
Gence-Creux, Jean-Michel Glachant, Tomás Gómez San Román, Leigh Hancher, Peter
Kaderjak, Christopher Jones, François Lévêque, Karsten Neuhoff, Ignacio Perez-Arriaga,
Andris Piebalgs, Michael Pollitt, Joe Perkins, Alberto Pototschnig, Konrad Purchala, Pippo
Ranci, Vincent Rious, Fabien Roques, Marcelo Saguan, Matti Supponen, Frauke Thies,
Leen Vandezande, Jorge Vasconcelos, Jean-Arnold Vinois, Niels-Henrik von der Fehr and
Christian von Hirschhausen.
This book has also benefited greatly from the online learning community at the Florence
School of Regulation. Thanks go to the directors and founders of this unique school:
Jean-Michel Glachant, Pippo Ranci, Ignacio Pérez-Arriaga and Jorge Vasconcelos. Thanks
also go to Klaus-Dieter Borchardt, Susanne Nies, Alberto Pototschnig and Laurent Schmitt for
supporting the training collaboration between the European Commission, ACER, ENTSO-E
and the Florence School of Regulation on the evolution of electricity markets with the EU
electricity network codes and the EU Clean Energy Package.
I thank the experts and colleagues who contributed to the online training sessions: Olivier
Aine, Daniela Bernardo, Klaus-Dieter Borchardt, Chiara Canestrini, Victor Charbonnier, Anna
Colucci, Paul de Wit, Ellen Diskin, Christian Dobelke, Edwin Edelenbos, Uros Gabrijel, Ilaria
Galimberti, Christophe Gence-Creux, Paul Giesbertz, Sotiris Georgiopolous, Leigh Hancher,
Christopher Jones, Anne-Marie Kehoe, Oliver Koch, Mathilde Lallemand, Christine Lyon,
Maria Eugenia Leoz Martin-Casallo, Claudio Marcantonini, Rafael Muruais Garcia, Rickard
Nilson, Martin Povh, Konrad Purchala, Mario Ragwitz, Ariana Ramos, Ioannis Retsoulis,
Hélène Robaye, Josh Roberts, Julius Rumpf, Manuel Sanchez-Jimenez, Peter Schell, Laurent
Schmitt, Salla Sissonen, Markela Stamati, Matti Supponen, Frauke Thies, Ernst Tremmel,
Sonya Twohig, Mark van Stiphout, Chara Vlachou, Kirsten Wilkeshuis and Annika Zorn.
I also thank the alumni and colleagues who helped us improve the course texts that have
inspired this book: Erik Ahlström, Olivier Aine, Eblerta Ajeti, Alia Al-Hathloul, Elena
Alonso, Christopher Andrey, Carlos Arsuaga, Adrien Atayi, Guénolé Aumont, Ross Baldick,
Pradyumna Bhagwat, Irina Boiarchuk, Joanna Bolesta, Anatolie Boscaneanu, Basile Bouquet,
Marco Campagna, Zaza Chikhradze, Daniel Daví-Arderius, Anne de Geeter, Miguel Manuel

xiii
xiv Evolution of electricity markets in Europe

de Villena, Francis Dike, Ellen Diskin, Piero Dos Reis, Žilvinas Dragūnas, Alexander Dusolt,
Gianluca Flego, Marco Foresti, Lars Olav Fosse, Mathieu Fransen, Eivind Gamme, Stamatia
Gkiala Fikari, Samson Hadush, Bastian Henze, Lenka Jancova, Tata Katamadze, Nico Keyaerts,
Mathias Katz, Vrahimis Koutsoloukas, Randi Kristiansen, Ketevan Kukava, Mariam Kukava,
Patrícia Lages, Mathilde Lallemand, Olivier Lamquet, Sam Lansink, Mathias Lorenz, Dijana
Martinčić, Emilie Milin, Matteo Moraschi, Nadine Mounir, Frank Nobel, Louise Nørring,
Marco Savino Pasquadibisceglie, Marco Pavesi, Konstantinos Petsinis, Ralph Pfeiffer, Martin
Pistora, Marta Poncela, Lorenz Rentsch, Ioannis Retsoulis, Martin Roach, Nicolò Rossetto,
Ali Sefear, Susana Serôdio, Ioannis Theologitis, Athanasios Troupakis, Per Arne Vada, Mihai
Valcan, Ellen Valkenborgs, Waldo Vandendriessche, Marco Vilardo, Michael Wilch, Steve
Wilkin, Nynke Willemsen, Peter Willis, Cherry Yuen and David Ziegler.
I would also like to thank the hundreds of participants who interacted with us in these online
training sessions and motivated us to write this book.
Finally, I thank Athir Nouicer, Valerie Reif and Tim Schittekatte. We co-teach in the online
training sessions and we also co-authored this book. They are the next generation of giants
in the field of electricity markets. We enjoyed writing this book and we hope you will enjoy
reading it.
Leonardo
List of abbreviations
4MMC 4M Market Coupling
AC Alternating Current
ACER (European Union) Agency for the Cooperation of Energy
Regulators
aFRP Automatic Frequency Restoration Process
aFRR Automatic Frequency Restoration Reserves
AM Available Margin
APX Amsterdam Power Exchange
Belpex Belgian Power Exchange
BRP Balance Responsible Party
BSP Balancing Service Provider
CACM GL Capacity Allocation and Congestion Management Guideline
CBA Cost–Benefit Analysis
CBCA Cross-Border Cost Allocation
CCGT Combined Cycle Gas Turbine
C-component Consumer component
CCR Capacity Calculation Region
CE Continental Europe
CEC Citizen Energy Community
CEE Central Eastern Europe
CEER Council for European Energy Regulators
CEF Connecting Europe Facility
CEP Clean Energy for all Europeans Package (Clean Energy Package)
CGM Common Grid Model
CGS Critical Grid Situation
CHP Combined Heat and Power
CNC Connection Network Code
CNE Critical Network Element
CO2 Carbon Dioxide
CoBA Coordinated Balancing Area

xv
xvi Evolution of electricity markets in Europe

CWE Central Western Europe

DC Direct Current
DC NC Demand Connection Network Code
DE Germany
DER Distributed Energy Resource
DG COMP Directorate-General for Competition
DLMP Distribution Locational Marginal Pricing
DSO Distribution System Operator
EB GL Electricity Balancing Guideline
EC European Commission
ECJ European Court of Justice
EENS Expected Energy Not Served
EEX European Energy Exchange
ENTSO-E European Network of Transmission System Operators for
Electricity
ENTSOG European Network of Transmission System Operators for Gas
EPAD Electricity Price Area Differential
ER NC Emergency and Restoration Network Code
ERGEG European Regulators’ Group for Electricity and Gas
ETSO European Transmission System Operators
EU European Union
EUPHEMIA Pan-European Hybrid Electricity Market Integration Algorithm
EXPLORE European X-border Project for LOng-term Real-time balancing
Electricity market design
FBMC Flow-Based Market Coupling
FCA GL Forward Capacity Allocation Guideline
FCR Frequency Containment Reserves
FERC Federal Energy Regulatory Commission
FRCE Frequency Restoration Control Error
FRP Frequency Restoration Process
FRR Frequency Restoration Reserves
FTR Financial Transmission Right
g gram
GB Great Britain
G-component Generator component
GCT Gate Closure Time
 Abbreviations xvii

GLDPM Generation and Load Data Provision Methodology

GLIP Gas Interconnection Poland–Lithuania
GSK Generation Shift Key
HVDC High Voltage Direct Current
HVDC NC Requirements for Grid Connection of High Voltage Direct
Current Systems and Direct Current-Connected Power Park
Modules Network Code
Hz Hertz
IGCC International Grid Control Cooperation
IGD Implementation Guidance Document
IGM Individual Grid Model
ISO Independent System Operator
ISP Imbalance Settlement Period
ITC Inter-TSO compensation
JAO Joint Allocation Office
kg kilogram
KORRR Key Organizational Requirements, Roles and Responsibilities
kWe kilowatt-electric
kWh kilowatt-hour
LFC Load Frequency Control
LOLE Loss of Load Expectation
LPX Leipzig Power Exchange
LSK Load Shift Key
MARI Manually Activated Reserves Initiative
MCO Market Coupling Operator
mFRR Manual Frequency Restoration Reserves
mHz millihertz
MNA Multi-NEMO Arrangement
MRC Multi-Regional Coupling
MW Megawatt
MWh Megawatt-hour
NEMO Nominated Electricity Market Operator
NORDEL Nordic Cooperation of Electricity Utilities
NRA National Regulatory Authority
NTC Net Transfer Capacity
OPDE Operational Planning Data Environment
xviii Evolution of electricity markets in Europe

OPPM Offshore Power Park Module

ORDC Operating Reserve Demand Curves
OTC Over the Counter
PCI Projects of Common Interest
PCN Physical Communication Network
PGM Power-Generating Module
PICASSO Platform for the International Coordination of Automated
Frequency Restoration and Stable System Operation
PPM Power Park Module
PTDF Power Transfer Distribution Factor
PTR Physical Transmission Right
PUN Prezzo Unico Nazionale
PV Photovoltaic
RAM Remaining Available Margin
RCC Regional Coordination Centre
REC Renewable Energy Community
RED II Directive (EU) 2018/2001 on the promotion of the use of energy
from renewable sources
RfG NC Requirements for Grid Connection of Generators Network Code
RoCoF Rate of Change of Frequency
RR Replacement Reserves
RSC Regional Security Coordinator
RSCI Regional Security Coordination Initiative
SCC Security Coordination Centre
SDAC Single Day-Ahead Coupling
SEE South-Eastern Europe
SGU Significant Grid User
SIDC Single Intraday Coupling
SMM Serbia, Macedonia and Montenegro
SO GL System Operation Guideline
SOR System Operation Region
SPGM Synchronous Power-Generating Module
SvK Svenska Kraftnät
TCM Terms and Conditions or Methodologies
TEN-E Trans-European Energy Networks
TERRE Trans-European Restoration Reserves Exchange
 Abbreviations xix

TSO Transmission System Operator

TYNDP Ten-Year Network Development Plan
UCTE Union for the Coordination of the Transmission of Electricity
UIOSI Use-It-Or-Sell-It
UK United Kingdom
US United States
VoLL Value of Lost Load
XBID Cross-Border Intraday
Introduction
In Europe, there is a sequence of electricity markets that starts years before the actual delivery
takes place and continues up to real time. Thanks to a regulatory process driven by European
Union institutions, most countries now have a very similar sequence of electricity markets.
Some of these markets have already evolved from national markets into European markets,
while others remain national or regional.
Figure 0.1 illustrates the sequence of electricity markets in Europe. There are wholesale
markets to price the electric energy commodity. They include forward markets as well as
day-ahead and intraday markets. There are also markets to put prices on transmission capacity.
These markets are integrated with the commodity markets for the short term, but not for the
longer term (Chapter 2). There are also separate markets for reserves or balancing services,
from year-ahead balancing capacity markets to close to real-time balancing energy markets
(Chapter 5). Finally, there are markets to correct the outcomes of the wholesale and balancing
markets.
Corrections are needed because wholesale and balancing markets do not yet fully consider
transmission and distribution network constraints. The corrective markets are referred to as
redispatching markets in the context of transmission networks (Chapter 3 and Chapter 5) and
flexibility markets in the context of distribution networks (Chapter 5 and Chapter 8). They

Figure 0.1 The sequence of electricity markets in Europe

xx
 Introduction xxi

can be separate markets or integrated into balancing markets or intraday markets. Corrections
are also needed if wholesale and balancing markets do not result in adequate investment in
generation capacity, demand-side flexibility or energy storage assets, and these corrections are
called capacity mechanisms. They price resource adequacy (Chapter 7).
Day-ahead markets were the first to be integrated into a European market, and were
followed by intraday and balancing energy markets. Forward transmission markets are cen-
tralized in one European platform (Chapter 2). Balancing capacity markets are still mainly
national or regional with only some aspects Europeanized. There is little effort to integrate
capacity mechanisms, and we do not yet know what will happen to redispatching markets and
flexibility markets. Finally, transit charges for transporting electricity over borders have been
abolished, and this happened long before roaming charges for mobile phones were abolished
(Chapter 4).
The benefits of electricity market integration were not clear at the start of the process.
Only after the markets developed at the national level with increasing transparency could the
benefits of integration at the European scale be assessed. Today, we have the evidence, and it
confirms that the benefits are significant (Chapter 1).
In this book, we tell the story of how Europeans have experienced the evolution of electric-
ity markets since the end of the 1990s. As far as we know, this experiment has not yet been
bundled into a book that bridges theory and practice. This has motivated us to document this
unique European creation.
Defining moments in the process include the four waves of European legislative pack-
ages (Chapter 1), a landmark court case and a software bug (Chapter 2), an Italian blackout
(Chapter 3), delayed clocks (Chapter 5), a cruise ship (Chapter 6), the financial crisis (Chapter
7) and school strikes and climate marches (Chapter 8).
In each chapter we will also refer to both issues that have been settled and issues that are still
open and we continue to face today. There are too many to list them all here, but here are a few
examples. Settled issues include explicit versus implicit auctions (Chapter 2), flow-based
market coupling (Chapter 3), the inter-TSO compensation mechanism (Chapter 4), single
versus dual pricing and regional versus European platforms in balancing markets (Chapter
5), and the connection requirements for wind and solar units (Chapter 6). Open issues include
product definitions in forward transmission markets and transmission allocation in the intraday
market (Chapter 2), the configuration of bidding zones (Chapter 3), creating a level playing
field for all grid users to access electricity markets (Chapter 4), the filtering of balancing bids
by system operators and transmission capacity reservation for balancing (Chapter 5), the con-
nection requirements for storage units (Chapter 6), the role of capacity mechanisms (Chapter
7), and a new paradigm to put the citizen at the centre of the energy transition (Chapter 8).
Looking back, most of the market integration process so far required horizontal coordi-
nation between transmission system operators from different countries. Looking forward,
vertical coordination between transmission and distribution system operators is becoming
more important, and this coordination process has only just started. We no longer talk about
completing the integration of electricity markets in Europe; instead, we embrace the process
and go with the flow. We hope you will do the same after reading this book.
If you are a European academic or practitioner, this book will provide you with the nec-
essary background to the debates that you are involved in. You will learn the language used
in the world of electricity markets in Europe. In each chapter, we include endnotes with the
xxii Evolution of electricity markets in Europe

main references and an annex with a regulatory guide to the legislative texts and regulatory
decisions that are driving the market integration process. A few key technical concepts are also
explained in the annexes.
If you are a non-European involved in the development of an international or multi-region
electricity market, the book gives you a list of issues you will face, together with the European
solutions. Your context is different, so you might need different solutions, but you can still
learn from the trial and error process we have gone through in Europe.
In combination with this European tour, you might like a global tour and a more in-depth
discussion of electricity market theory. We warmly recommend that you read the Handbook
on Electricity Markets by Jean-Michel Glachant, Paul Joskow and Michael Pollitt, which
together with this book is published by Edward Elgar.
Structure of this book

–– PART I: HOW TO TRADE AND TRANSPORT ELECTRICITY ACROSS NATIONAL

BORDERS?
–– Chapter 1: Why did we start with electricity markets in Europe?
–– Chapter 2: Who gets the rights to trade across borders?
–– Chapter 3: How do you calculate border trade constraints?
–– Chapter 4: Who pays for the network when trade is international?

Electricity is not transported in trucks or by rail as it cannot easily be stored. It is delivered

instantly when we need it through the cables that connect our homes to a system that includes
many power plants that jointly produce the electricity we need when we use it. It is one of the
engineering wonders of our time. To allow electricity markets to function, we need many rules
and regulations that clarify the roles and responsibilities of all involved. European countries
have only been willing to harmonize these rules when it has been necessary to capture the
benefits from integrating markets across national borders. We have often only done the right
thing after trying everything else. In this first part of the book, we highlight the issues that we
have faced since the start of the integration of electricity markets in Europe.

–– PART II: HOW TO COMBINE ELECTRICITY TRADING WITH SYSTEM SECURITY

TO KEEP THE LIGHTS ON?
–– Chapter 5: Who is responsible for balancing the system?
–– Chapter 6: How to organize system operation and connection requirements?
–– Chapter 7: How to ensure adequate investment in power plants?

For other products and services, markets produce price signals that limit shortages. If a short-
age seems to be on the horizon, prices go up and investment follows. If shortages do occur,
the users of the product or service who are last in line have to wait. In the case of electricity,
if there are shortages there is rationing, so we share the pain. Candles at home can be cosy
for a few hours but they provide little consolation when you are trapped in a lift or a metro.
Without electricity, nothing in our society works. All kinds of safeguards have therefore been
put in place to intervene in electricity markets when needed. These interventions tend to be
more national than European, but some progress has been made in cross-border collaboration
in Europe to reduce costs and to limit market distortions. In the second part of the book, we
discuss how this works.
–– PART III: HOW TO PUT THE CITIZEN AT THE CENTRE OF THE ENERGY
TRANSITION?
–– Chapter 8: How to put the citizen at the centre of the energy transition?

xxiii
xxiv Evolution of electricity markets in Europe

We used to think of citizens as electricity consumers. When electricity markets were intro-
duced to replace the national monopolies that dominated the sector, the promise was that it
would result in cheaper prices and better service. Soon after markets were introduced, Europe
also started to decarbonize its electricity sector by subsidizing renewable energy and energy
efficiency. The energy transition made the system more sustainable but drove up prices, as
about a third of an electricity bill is made up with taxes and levies to recuperate costs from
subsidies that have been granted. To keep them on board, citizens have now been offered
a new deal. Instead of remaining passive consumers, they can become active by producing
their own electricity with smart buildings, by engaging in peer-to-peer trade or by joining
energy communities. In the third part of the book, we discuss these recent developments and
their impact on electricity markets.
PART I

How to trade and transport electricity across national

borders?
1. Why did we start with electricity markets in
Europe?
Leonardo Meeus with Valerie Reif

In this chapter we answer three questions. First, what was the political process that led to
electricity markets in Europe? Second, what were the technical drivers for creating a European
power system? Third, what do we know about the benefits of integrating electricity markets
in Europe?

1.1 WHAT WAS THE POLITICAL PROCESS THAT LED TO

ELECTRICITY MARKETS IN EUROPE?

From a political perspective, the integration of electricity markets followed three main steps,
with increasing levels of detail: European treaties, EU legislative energy packages and more
detailed market rules that have been developed in the process of creating EU electricity
network codes and guidelines.
First are the European treaties. The aim to create a common market to eliminate trade barri-
ers between Member States dates back to the founding Treaty of Rome in 1957. Twenty-nine
years later, the Single European Act of 1986 was adopted as the first major revision of the
Treaty of Rome. It paved the way for what was to become one of the main achievements of
the European project: the Single European Act required the adoption of measures with the
aim of establishing an internal market by 31 December 1992.1 Later that year, the Council
(1986) adopted energy policy objectives for the European Community, among which was that
of ‘greater integration, free from barriers to trade, of the internal energy market with a view
to improving security of supply, reducing costs and improving economic competitiveness’.
In 1988, the Commission of the European Communities published the first document on the
internal energy market, which assessed that there were still considerable barriers to trade in
energy products within the Community (EC 1988). On 1 January 1993, the European Single
Market became a reality for the 12 Member States at that time. However, integrating the
energy sector into the European Single Market alongside other goods proved lengthier and
more complex than had originally been anticipated. The year 1993 turned out to be only the
starting point of a long, and still ongoing, process to build an EU internal market for electricity,
as we will show in this book. One reason was the legacy structure of the energy sector and the
non-existence of markets at the national level. Up to the mid-1990s, the electricity sector was
still dominated by state-owned or state-controlled vertically integrated utilities with regional
or national monopolies. Cross-border trade was limited due to a lack of infrastructure and rules
to organize this trade.

2
 Why did we start with electricity markets in Europe?

Figure 1.1 The main steps in the evolution of European electricity markets
3
4 Evolution of electricity markets in Europe

Second are the EU legislative energy packages. Profound changes were introduced to the
national electricity sectors in three Electricity Market Directives in 1996, 2003 and 2009, as is
shown in Figure 1.1. More recently, a fourth directive was adopted in 2019. All the directives
are part of a so-called energy package. While the first three packages included one directive
each for the electricity and the gas sectors plus a varying number of regulations for both
sectors, the Clean Energy for All Europeans Package (Clean Energy Package, CEP) did not
address the gas sector directly. Curiously enough, the naming of these packages follows its
own logic, as is described in Box 1.1.

BOX 1.1 NICKNAMING EU LEGISLATION

Over the years, we have made an interesting observation regarding the naming of EU
directives and regulations in the electricity sector.
Almost all the pieces of legislation have received nicknames that are widely applied. As
a result of this practice, nobody remembers any of the official numbers of the regulations.
In other words, Directive 96/92/EC is known as the First Directive, Directive 2003/54/EC
as the Second Directive and so forth. One exception is Regulation 1228, where the number
became the nickname.
It should also be noted that the development of nicknames lacked consistency and logic.
Prior to the third package becoming the Third Package, it was called the Two Plus Package
for a short while. This was to imply that the resulting package would merely consist of
light amendments to the Second Package rather than a new package, which it turned out to
be in the end. In the case of the latest package, several nicknames existed, such as Fourth
Package, 2030 Package and Jumbo Package. After a brief period of Winter Package, the
winner that finally emerged was Clean Energy Package.
Perhaps you are wondering why we even mention these nicknames. The reason is that
if you want to become part of the industry the importance of speaking its language should
not be underestimated.

The First Package and Directive 96/92/EC (First Directive) kicked off the liberalization
process by introducing a distinction between the regulated part of the sector (network) and
the competitive parts (generation and supply). However, it left a large margin of choice for
the Member States as to how to introduce more competition into their electricity markets,
resulting in significant differences in the level of market opening. Despite its failure to deliver
the degree of liberalization originally intended, the First Directive gave the Member States
and their national utility champions a taste of what was to come.2 In the following decades,
national markets were gradually opened with the Second, the Third and the Clean Energy
Package entering into force. As we will refer to the changes brought by the CEP throughout
this book, Box 1.2 provides an introduction to it.
 Why did we start with electricity markets in Europe? 5

BOX 1.2 THE EU CLEAN ENERGY PACKAGE

The Clean Energy Package is the latest of four packages that have been introducing fun-
damental changes to national electricity sectors since the 1990s. The CEP development
process aimed to push forward the energy transition that started with the publication of
draft legislative texts by the European Commission in November 2016. In June 2019, the
adoption process was completed following the publication of the final legislative texts in
the Official Journal of the European Union. The CEP consists of four directives and four
regulations, which are listed below.
The regulations are Regulation (EU) 2018/1999 or ‘Regulation on the Governance
of the Energy Union’, published on 21 December 2018; Regulation (EU) 2019/941 or
‘Regulation on Risk-Preparedness’, published on 14 June 2019; Regulation (EU) 2019/942
or the ‘(Recast of the) ACER Regulation’, published on 14 June 2019; and Regulation (EU)
2019/943 or the ‘(Recast of the) Electricity Regulation’, published on 14 June 2019.
The directives are Directive (EU) 2018/844 or the ‘Energy Performance in Buildings
Directive’, published on 19 June 2018; Directive (EU) 2018/2001 or the ‘Renewable
Energy Directive (RED II)’, published on 21 December 2018; Directive (EU) 2018/2002
or the ‘Energy Efficiency Directive’, published on 21 December 2018; and Directive (EU)
2019/944 or the ‘(Recast of the) Electricity Directive’, published on 14 June 2019.
Regulation (EU) 2019/943 and Directive (EU) 2019/944 are the ones of most importance
in the developments described in this book. The date of application of Regulation (EU)
2019/943 is 1 January 2020. Member States have 18 months to transpose Directive (EU)
2019/944 into national law. The European Commission will review the implementation
of Directive (EU) 2019/944 and Regulation (EU) 2019/943 by 31 December 2025 and
31 December 2030 respectively. The Commission will submit a report to the European
Parliament and the Council, accompanied by legislative proposals where appropriate.

Note that a major part of the early legislation focused on setting the conditions and creating the
institutions necessary for electricity markets to function, but not on the actual market design
or detailed market rules. In what follows, we illustrate this for transmission system operators
(TSOs) and national regulatory authorities (NRAs).
As is illustrated in Figure 1.2, TSOs have been gradually made more independent from gen-
eration, which is referred to as the unbundling process. The First Package only required man-
agement and accounting unbundling. The Second Package took unbundling a step further in
requiring transmission and distribution companies to apply legal unbundling from 1 July 2004
and 1 July 2007 respectively. The Third Package finally provided that as of 3 March 2012
TSOs had to be certified by the competent NRA under one of the three unbundling models:
full Ownership Unbundling; Independent System Operator; and Independent Transmission
Operator. Ownership unbundling emerged as the dominant model in Europe.3 The Third
Package also required all the TSOs to create the European Network of Transmission System
Operators for Electricity (ENTSO-E) and to cooperate through this new institution at the
European level. Figure 1.2 lists some of the TSOs’ and ENTSO-E’s tasks, which have clearly
been increasing with each legislative package that has been adopted. We do not discuss them
here, as they will be covered in the various chapters of this book. The Clean Energy Package
also requires the establishment of an entity of distribution system operators in the Union (EU
 6
Evolution of electricity markets in Europe

Figure 1.2 The development of TSOs at the national and European levels and a selection of their tasks
 Why did we start with electricity markets in Europe? 7

Figure 1.3 The development of regulatory authorities at the national and European
levels and a selection of their tasks

DSO entity) to increase efficiencies in the electricity distribution networks and to ensure close
cooperation with TSOs and ENTSO-E.4
As is illustrated in Figure 1.3, NRAs have gradually been made more independent of the
industry and national governments. Initially, some countries like Germany did not see the need
for an energy regulator but relied on combinations of self-regulation and competition author-
ities. The Second Package eventually put an end to such arrangements by requiring Member
States to create national regulatory bodies that are independent of the electricity industry. The
Third Package increased the independence of NRAs from national governments and also man-
dated the establishment of the Agency for the Cooperation of Energy Regulators (ACER).5
Figure 1.3 lists some of the NRAs’ and ACER’s tasks, which have clearly been increasing with
each legislative package that has been adopted. We do not discuss them here, as they will be
covered in the various chapters of this book.
Third are the more detailed market rules that have been developed through the process
of creating EU network codes and guidelines. The first generation of network codes and
guidelines were adopted after a lengthy co-creation process involving the European institu-
tions, ENTSO-E, ACER and many stakeholders from across the electricity sector. This first
generation consisted of eight legislative acts that entered into force between 2015 and the end
of 2017. As we will refer to these network codes and guidelines throughout the book, Box 1.3
provides an introduction to the eight network codes and guidelines, their scope, and the related
development, implementation and amendment processes.
8 Evolution of electricity markets in Europe

BOX 1.3 THE FIRST GENERATION OF EU ELECTRICITY

NETWORK CODES AND GUIDELINES

The network codes and guidelines of the first generation can be subdivided into three
groups or ‘code families’ as listed below.

• Market codes: the capacity allocation and congestion management guideline (CACM
GL), published on 25 July 2015; the forward capacity allocation guideline (FCA GL),
published on 27 September 2016; the electricity balancing guideline (EB GL), pub-
lished on 23 November 2017.
• Connection network codes (CNCs): the network code on requirements for grid connec-
tion of generators (RfG NC), published on 14 April 2016; the network code on demand
connection (DC NC), published on 18 August 2016; the network code on requirements
for grid connection of high voltage direct current systems and direct current-connected
power park modules (HVDC NC), published on 8 September 2016.
• Operation codes: the electricity transmission system operation guideline (SO GL), pub-
lished on 25 August 2017; the electricity emergency and restoration network code (ER
NC), published on 24 November 2017.

These are commonly referred to as ‘the network codes’, but not all of them are legally de-
fined as network codes. Four of the eight are guidelines (CACM GL, FCA GL, EB GL and
SO GL) and the other four are network codes (ER NC, RfG NC, DC NC and HVDC NC).
Initially, all eight were planned to be developed as network codes, yet some became guide-
lines in the development process. In theory, network codes and guidelines can cover the
same topics. In practice, however, it is observed that some topics lend themselves better to
guidelines than to network codes and others vice versa.
Both similarities and differences between network codes and guidelines exist. Network
codes and guidelines are similar in that they carry the same legal weight (both are
Commission regulations and are legally binding), are directly applicable (they do not need
to be transposed into national law) and are subject to the same formal adoption procedure
(‘old’ comitology procedure). Network codes and guidelines differ from each other regard-
ing their legal basis, the stakeholder involvement, their amendment process, topics and
scope, and the adoption of further rules during the implementation phase. Indeed, the main
difference is the work to be done during the implementation phase, which we explain in the
following.
In general, network codes are more detailed than guidelines. Guidelines shift a larger
share of the further development to the implementation phase, which can allow for more
flexibility but can also slow down or complicate the overall process. Guidelines include
processes whereby TSOs or Nominated Electricity Market Operators (NEMOs, see also
Chapter 2) must develop so-called ‘Terms and Conditions or Methodologies (TCM)’. TCMs
are comprehensive (legal) texts that are often referred to as ‘methodologies’. In most cases,
methodologies have to be jointly developed by all TSOs or all NEMOs at the pan-European
level or by the relevant TSOs/NEMOs at the regional or national levels. Depending on the
scope of the methodologies, the Third Package foresaw their approval either by all NRAs
(pan-European methodologies) or by the relevant subset of NRAs (regional and national
 Why did we start with electricity markets in Europe? 9

methodologies). In certain cases, a decision is to be referred to ACER. ENTSO-E, ACER

and the European Commission have monitoring, reporting and stakeholder involvement
responsibilities related to methodologies. The implementation of the TCMs foreseen in the
first generation of network codes and guidelines will continue until around 2025. We will
refer to some of these methodologies in this book where relevant.
The regulatory guide in Annex 1A.1 provides detailed references to the relevant
legislation.

Note: Our colleagues Hancher et al. (2020) published a research report which provides more details on the legal
technicalities of the EU electricity network codes and guidelines.

For the first generation of network codes and guidelines, TSOs were placed in the position of
drafting network codes through ENTSO-E with regulatory oversight. You may wonder why
regulators have not been put in the position to develop these market rules. The typical answer
given to this question in the European context is that the level of detail and technical com-
plexity was such that the industry was asked to develop solutions that were then challenged
by the regulators rather than vice versa. However, the perception has been that TSOs have not
always developed solutions fast enough and that stakeholder involvement in the process has
been insufficient. Following the Clean Energy Package, significant changes including shifts in
roles and responsibilities have been introduced for both existing and future generations of EU
network codes and guidelines, as is shown in Box 1.4.

BOX 1.4 IMPLICATIONS OF THE EU CLEAN ENERGY

PACKAGE FOR EU NETWORK CODES AND
GUIDELINES

In 2019, the adoption of the Clean Energy Package brought significant changes for both
existing and future generations of EU network codes and guidelines. First, the development
process saw a shift in roles and responsibilities. The strong role of ENTSO-E in drafting the
network codes was reduced. The CEP also mandates the establishment of an EU DSO entity
to involve distribution system operators (DSOs) in the network code and guideline draft-
ing process. The role of ACER in the development phase is expected to increase. Another
change concerns the time interval in which the European Commission is required to com-
pile a priority list for new network codes.
Second, changes were introduced to the adoption process for both the TCM and new net-
work codes and guidelines. Regarding TCMs, ACER now directly decides on the method-
ologies with a pan-European scale (former ‘all NRA’ decisions). Regarding network codes
and guidelines, the Clean Energy Package distinguishes between the adoption of network
codes and guidelines as implementing or delegated acts. Depending on the type of act, the
European institutions and stakeholders have different rights and possibilities to intervene
in the adoption process.
In other words, the story of EU network codes and guidelines as a way to push forward
the market integration process in Europe continues. The scope of areas in which detailed
market rules can be developed has increased and the process has been fine-tuned. How it
will work in detail remains to be seen and can be reported in a future edition of this book.
10 Evolution of electricity markets in Europe

Note that we now have thousands of pages filled with detailed market rules, but they do not
prescribe a standard market design that everybody needs to follow. In fact, the new rules
simply reduce the degree of freedom that countries have in designing their markets. This is
why it is difficult to read the European market rules. They do not explicitly explain the design;
the resulting design is implicit. In this book, we will make it explicit.

1.2 WHAT WERE THE TECHNICAL DRIVERS FOR CREATING

A EUROPEAN POWER SYSTEM?

In this section we will see that there are technical advantages of scale.6 For example, a larger
power system is more stable as it has a higher level of inertia, which makes it easier for system
operators to keep the lights on. Another technical advantage of integration in this context was
the development of a solidarity mechanism between European countries that pre-dated the
creation of markets. To better understand these technical issues, we will use a tandem bicycle
analogy that is often used to explain power systems to non-engineers. Note that we are not
exhaustive in our description and only refer to certain elements of this analogy.7
Imagine a tandem bicycle moving at a constant speed as illustrated in Figure 1.4. The task
of the dark grey cyclists (power stations) is to generate the electrical energy that keeps the
entire system going. The light grey figures (loads) are not generating any of this energy, but the
aim of the overall system is to keep them moving nevertheless. The chain connecting all the
elements in the system represents the electrical transmission network. To maintain the same
velocity the chain must turn the wheels at a constant rate. In addition, constant physical tension
is required in the upper part of the chain. These two features can be respectively translated into
a need for a fixed constant frequency to guarantee a well-functioning system and a need for
a fixed voltage level for grid connections in an electricity network.
To transmit the pedalling movement (energy) to the chain, different connections between
the cyclists and the chain exist. A first type of dark grey cyclist (large thermal and nuclear
power stations connected to the grid with transformers) have their pedals directly connected to
the chain with one gear, which means they have to constantly pedal at the right speed and with
the right amount of power. A second type of dark grey cyclist (e.g. hydropower stations with

Note: Load is light grey and generation is dark grey.

Figure 1.4 A basic representation of the power system using a tandem bicycle
 Why did we start with electricity markets in Europe? 11

their turbines connected to generators) may prefer to cycle more slowly and have their force
transformed to the right speed with a gear system. A third type (e.g. wind turbines connected
with a frequency inverter) is connected through a belt and a gear system, allowing them to
pedal at varying speed. There are also different types of loads, but we will not explain them
in detail here.
The complex task of power system management requires both the speed (frequency) and the
tension of the chain (voltage level) to remain steady, even in the event of unexpected imbal-
ances between load and generation. One of the technical advantages of scale is that the more
synchronously connected rotating machines a power system has (dark grey cyclists types one
and two), the more stable it is as it has a higher level of inertia. Inertia represents the ability
of synchronously connected rotating machines to store and inject their kinetic energy into
the system. Inertia slows down a frequency drop/spike immediately after a sudden mismatch
between supply and demand (e.g. a power station outage or an unexpected change in the load
connected to the network). If the system inertia is low, a small sudden difference between
load and generation causes a high-frequency deviation. It is important to note that inertia
only supports frequency in nearly instantaneous situations where an imbalance is caused by
a sudden disconnection of large units or a nearly instantaneous change in production or load.
Inertia does not support frequency under ‘normal’ imbalance conditions when the imbalance
is caused by a prognosis error and resulting differences between production and consumption
plans. System inertia is typically higher for larger synchronous power systems as the kinetic
energy available and therefore the system’s inertia increases with the number of generators
and motors that are coupled to the grid.
However, system inertia can only slow down frequency deviations; it is not able to restore
the power balance between generation and load. Therefore, some of the dark grey cyclists
(power stations) do not pedal at full power. Instead, they conserve some of their energy to
be able to provide extra or replacement force (reserves) when it is needed. TSOs, which are
responsible for safeguarding system security within their control areas, must ensure that there
are enough reserves to regulate frequency and respond to possible emergency situations. In
a stand-alone system, such reserves must typically be large enough to cope with the most
severe incident, which usually corresponds to a loss of the largest generator in one TSO’s
control area. A clear advantage of interconnecting control zones to form a large synchronous
area is that reserves can be pooled and the relative importance of the most severe incident
decreases as system size increases. Such solidarity schemes for reserve sharing in which
each TSO can draw on the reserves in other TSOs’ control zones whenever needed were
implemented in synchronous areas long before markets were introduced. We will come back
to the balancing mechanism that is in place today in Chapter 5. We will also come back to the
technical requirements that different types of assets that are connected to the power system
need to comply with in Chapter 6.

1.3 WHAT DO WE KNOW ABOUT THE BENEFITS OF

INTEGRATING ELECTRICITY MARKETS IN EUROPE?

The initial focus of cross-border cooperation was on system stability as mentioned in the pre-
vious section and sharing of reserves as we will explain in the following. Sharing of reserves
aimed at a more effective use of energy resources and optimal operation of electric power
12 Evolution of electricity markets in Europe

plants enabled by the interconnection of electricity networks. For example, by exporting

electricity across borders, countries with large hydropower resources could prevent surplus
production in their own country while at the same time allowing for savings in coal consump-
tion in the neighbouring country. Similar qualitative reasoning is still valid in Europe today,
as a single energy market is believed to take better advantage of the differences in industrial
policy, available natural resources, weather conditions and load patterns across countries. For
example, France is dominated by nuclear, Germany still has a large share of coal and Norway
has significant hydro capacities, while the UK, Spain and Italy have higher relative shares of
gas in their electricity mix. Countries with a large share of electrical heating such as France
and Norway have larger differences between summer and winter consumption levels than
countries with other heating sources. Time differences between countries can also be a source
of differences in demand peaks. More recently, an important aspect of integration has been
a better use of renewable resources due to uneven wind and solar conditions across Member
States. Trade between European countries came only later, which was also due to the structure
of the energy sector dominated by vertically integrated monopolies. However, already in the
pre-liberalization period, long-term cross-border power purchase agreements with neighbour-
ing utilities were sometimes preferred to authorizing the construction of new domestic power
plants.
In the early 2000s the new competitive environment was slowly manifesting itself as
a result of the first two energy packages. At that time, questions were raised on how large the
benefits of liberalization were, how these benefits could be reaped, and who could reap these
benefits. The European Commission published several benchmarking reports to evaluate the
implementation of electricity (and gas) directives in the early 2000s. These reports put forward
indicators to check the health of particular markets. Competitive activity was measured with
market development indicators related to concentration and new entries (e.g. the share of
the three biggest generators, the share of the three biggest suppliers, the main retail supplier
entrant type), switching estimates for different types of customers, price development (price
convergence between Member States, price levels for different customer groups) and trade
between Member States (the level of cross-border electricity exchange, use of interconnector
capacity). None of the reports, however, included a quantification of the benefits of integrating
national electricity markets to create an EU internal market for electricity. Only around 2012
did the European Commission task consultants with assessing the benefits of an integrated
European electricity market. They concluded that the benefit was several billions of euros
a year, an order of magnitude that was later confirmed by ACER. In its annual market moni-
toring report, ACER has been gradually improving its methodology to estimate these benefits
and has also gone a long way in making the information that goes into the calculation more
readily available. The ACER annual market monitoring report is a must-read for everybody
reading this book. As it is highly technical, this book will give you the necessary background
to be able to read it.8
More recently, the political developments around Brexit have led academics to estimate the
potential cost of disruptions in electricity trade if the UK were to leave the European Union
without a suitable trade deal. The electricity market in Great Britain, that is England, Scotland
and Wales, is currently a single bidding zone, which is connected to Belgium, France and the
Netherlands, and also to Northern Ireland and the Republic of Ireland. At the time of writing,
uncertainty remains as to whether the political Brexit includes an ‘Elecxit’ (the UK leaving
 Why did we start with electricity markets in Europe? 13

the EU internal energy market) and what longer-term consequences such a loss of a relatively
small piece of the European electricity markets and the halting of interconnector expansion
between Great Britain and neighbouring EU Member States would have. Academics have
found that the 2030 cost to Britain of a hard electricity Brexit with little interconnector expan-
sion and decoupled markets would amount to several hundreds of million euros a year.9

1.4 CONCLUSION

In this first chapter on why we started with electricity markets, we have answered three ques-
tions. First, what was the political process that led to electricity markets in Europe? The aim
of creating a European internal energy market had been on the agenda of the European institu-
tions since the 1980s. Over the period 1996 to 2019, four EU legislative energy packages were
adopted that aimed to integrate and harmonize national electricity markets and mandated the
creation of ENTSO-E and ACER to drive the process. EU network codes and guidelines were
created that set out more detailed market rules. The Clean Energy Package introduced signif-
icant changes to the processes of developing, adopting, amending and implementing existing
and future generations of network codes and guidelines.
Second, what were the technical drivers for creating a European power system? A clear
advantage of larger power systems is their greater stability as they typically have larger
numbers of synchronously connected rotating machines which increases the level of system
inertia. Another technical advantage of scale is that the relative importance of the most severe
incident decreases as system size increases and reserves to cope with such incidents can be
pooled across TSO control zones. Such technical cooperation and reserve sharing among
TSOs for mutual support through interconnections in case of emergencies pre-dated markets.
Third, what do we know about the benefits of integrating electricity markets in Europe? The
qualitative benefits of exchanging electricity across borders to take advantage of the differ-
ences in generation mixes, weather conditions and load patterns have long been acknowledged
by European countries. A more recent motivation has been to make better use of renewable
resources due to uneven wind and solar conditions across Member States. The economics of
market integration became clearer when information became available to assess the benefits.
In 2013, an initial study found that the benefit of implementing an integrated EU electricity
market was several billions of euros a year, an order of magnitude that was later confirmed by
ACER. More recently, the potential cost to Great Britain of leaving the internal energy market
as a part of the political Brexit has been estimated to be in the order of several hundred million
euros a year by 2030.

NOTES
1. It can be confusing to keep up with the varying terminology used for the European Single Market
project, that is, common, internal and single market. The original treaties used the term ‘common
market’ without providing a definition. Legal literature suggests that the concept of the common
market went beyond the four freedoms and also included various policy areas such as agriculture,
competition and state aid. Later, the term ‘common market’ was replaced with ‘internal market’ in
the treaties, referring to an area without internal frontiers in which the free movements of goods,
persons, services and capital are ensured. While the objectives remained the same, the procedures
to adopt related legislation changed from unanimity (common market) to qualified majority voting
14 Evolution of electricity markets in Europe

(internal market). The term ‘single market’ can be considered an informal synonym of ‘internal
market’. Note that in some languages, only one word exists (e.g. ‘Binnenmarkt’ in German).
2. Hancher (2002) discusses the successes and failures of the First Electricity Directive of 1996 in
introducing a competitive environment in national electricity sectors. Vasconcelos (2005), the
founder of the Council for European Energy Regulators (CEER) in 2000, explains that the First
Energy Directive provided little guidance as regards cross-border energy trade, the development
of regional markets, interaction with non-EU markets, the development of interconnectors, the
supra-national integration of energy markets and so on. Hence, a ‘regulatory gap’ between national
markets and the EU internal energy market emerged. He elaborates on how Regulation (EC) No
1228/2003, which we will discuss in Chapter 2, represented the Commission’s attempt to close the
‘regulatory gap’, which was shown to be not possible on a voluntary basis.
3. CEER (2016) provides an overview and explanation of the different unbundling models applied
across European Member States. TSOs also continue to change, as is seen in the example of the
Greek TSO. ADMIE moved from the Independent Transmission Operator model to the Ownership
Unbundling model as a consequence of changes in the ownership and share structure (CEER 2019).
4. We do not cover the EU DSO entity in this book as there are still many open issues at the time of
writing. In the future, the EU DSO entity is expected to play a significant role for example as regards
the preparation and implementation of new network codes, where relevant for distribution networks.
We might therefore include its tasks and responsibilities in a future edition of this book.
5. Glachant et al. (2008) analyse the institutional mechanisms of the German self-regulation arrange-
ment that lasted from 1998 to 2005. Pototschnig (2019) provides a comprehensive overview of
developments from the creation of ACER to its future as foreseen in the Clean Energy Package.
Jones (2016) provides a full overview of the Third Energy Package, including the topics discussed
here of common electricity wholesale markets, the unbundling of TSOs, NRAs and the coming into
existence of ACER, and the regulation of cross-border electricity exchanges.
6. Many of the technical, economic and regulatory fundamentals we touch on in this book are dis-
cussed in depth in Pérez-Arriaga (2013), which is a must-read for anybody entering the sector or
wanting to refresh their knowledge of some of the basic concepts the electricity sector deals with on
a daily basis.
7. We do not know who came up with this analogy, but were inspired by Söder (2002) and Fassbinder
and De Wachter (2005). Many thanks to our colleague Daniela Bernardo, who produced the
drawing in Figure 1.4, which is an enhanced version of a drawing in Söder (2002).
8. DG TREN (2001) was the first of four benchmarking reports published by the Commission of the
European Communities. Booz & Company (2013) is the consultancy study. ACER and CEER
(2019) is the annual market monitoring report. Newbery et al. (2016) provide an academic discus-
sion of the methodology that is used in these reports.
9. This paragraph is based on the work of Geske et al. (2020), who calculate the estimated 2030 cost
of Great Britain leaving the EU internal market for electricity based on a microeconomic model of
decoupled markets between Great Britain and France. Newbery (2020) discusses their results in
a brief review for Nature Energy.

REFERENCES
ACER and CEER (2019), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2018 – Electricity Wholesale Markets Volume’.
Booz & Company (2013), ‘Benefits of an Integrated European Energy Market’, Final Report prepared
for Directorate-General Energy European Commission.
CEER (2016), ‘Status Review on the Implementation of Transmission System Operators’ Unbundling
Provisions of the 3rd Energy Package’, CEER Status Review Ref: C15-LTF-43-04.
CEER (2019), ‘Implementation of TSO and DSO Unbundling Provisions – Update and Clean Energy
Package Outlook’, CEER Status Review Ref: C18-LAC-02-08.
Council (1986), ‘Council Resolution of 16 September 1986 Concerning New Community Energy Policy
Objectives for 1995 and Convergence of the Policies of the Member States (86/C 241/01)’.
 Why did we start with electricity markets in Europe? 15

DG TREN (2001), ‘First benchmarking report on the implementation of the internal electricity and gas
market. SEC (2001) 1957’, Commission Staff Working Paper.
EC (1988), ‘The internal energy market’, Commission Working Document COM(88) 238 Final.
Fassbinder, S. and B. De Wachter (2005), ‘The electrical system as a tandem bicycle’, accessed at www​
.gonder​.org​.tr/​wp​-content/​uploads/​2015/​04/​ElectricityTandem​.pdf​%0A​%0A.
Geske, J., R. Green and I. Staffell (2020), ‘Elecxit: The cost of bilaterally uncoupling British–EU elec-
tricity trade’, Energy Economics, 85, 104599.
Glachant, J.-M., U. Dubois and Y. Perez (2008), ‘Deregulating with no regulator: Is the German electric-
ity transmission regime institutionally correct?’, Energy Policy, 36 (5), 1600–10.
Hancher, L. (2002), ‘Slow and not so sure: Europe’s long march to electricity market liberalization’,
Electricity Journal, 10 (9), 92–101.
Hancher, L., A.-M. Kehoe and J.Rumpf (2020), ‘The EU Electricity Network Codes and Guidelines:
A Legal Perspective’, Research Report, European University Institute.
Jones, C. (ed.) (2016), EU Energy Law Volume I: The Internal Energy Market, 4th ed., Deventer,
Netherlands and Leuven, Belgium: Claeys & Casteels.
Newbery, D. M. (2020), ‘The cost of uncoupling’, Nature Energy, 5 (3), 187–8.
Newbery, D., G. Strbac and I. Viehoff (2016), ‘The benefits of integrating European electricity markets’,
Energy Policy, 94, 253–63.
Pérez-Arriaga, I. J. (ed.) (2013), Regulation of the Power Sector, Springer-Verlag London.
Pototschnig, A. (2019), ‘The ACER experience’, in S. Nies (ed.), The European Energy Transition: Actors,
Factors, Sectors, Deventer, Netherlands and Leuven, Belgium: Claeys & Casteels, pp. 175–211.
Söder, L. (2002), ‘Explaining power system operation to non-engineers’, IEEE Power Engineering
Review, 22 (4), 25–7.
Vasconcelos, J. (2005), ‘Towards the internal energy market, how to bridge a regulatory gap and build
a regulatory framework’, European Review of Energy Markets, 1 (1).
16 Evolution of electricity markets in Europe

1A.1 ANNEX: REGULATORY GUIDE

Table 1A.1 Regulatory guide

Section of this chapter, topic and relevant regulation Relevant articles

Section 1.1

The aim to create a common market to eliminate trade barriers Art. 2 of the Treaty of Rome states that ‘The Community
between Member States dates back to the founding Treaty of shall have as its task, by establishing a common market and
Rome in 1957. progressively approximating the economic policies of Member
States, to promote throughout the Community a harmonious
development of economic activities, a continuous and balanced
expansion, an increase in stability, an accelerated raising of
the standard of living and closer relations between the States
belonging to it.’
The Treaty includes, among many other things, provisions on
the free movement of goods (Title I) and the free movement of
persons, services and capital (Title III).
The Single European Act of 1986 required the adoption of Art. 13 of the Single European Act states that ‘… the
measures with the aim of establishing an internal market by 31 Community shall adopt measures with the aim of progressively
December 1992. establishing the internal market over a period expiring on 31
December 1992 … The internal market shall comprise an area
without internal frontiers in which the free movement of goods,
persons, services and capital is ensured …’
The Council (1986) adopted energy objectives for the European In the Council Resolution of 16 September 1986, the Council
Community. of the European Communities ‘… 5. considers that the energy
policy of the Community and of the Member States must
endeavour to achieve the following horizontal objectives: … (d)
greater integration, free from barriers to trade, of the internal
energy market with a view to improving security of supply,
reducing costs and improving economic competitiveness.’
In 1988, the Commission of the European Communities The Commission Working Document COM(88) 238 final on
published the first document on the internal energy market, the internal energy market of 2 May 1988 states in its Part Two
which assessed that there were still considerable barriers to on the suggested priorities regarding the obstacles related to
trade in energy products within the Community. the establishment of a single energy market that the ‘barriers
are very diverse in type and significance … Most of them are
the end-product of domestic rules and regulations originating
in an often distant past predating European ideas: this applies
for example to all the potential obstacles arising from purely
domestic monopolies. …’
Directive 96/92/EC kicked off the liberalization process by As illustrations, recital 22 of Directive 96/92/EC states that ‘it
introducing a distinction between the regulated part of the is … necessary to establish common rules for the production
sector and competitive parts. of electricity and the operation of electricity transmission and
distribution systems’; and recital 30 states that ‘in order to
ensure transparency and non-discrimination, the transmission
function of vertically integrated undertakings should be
operated independently from the other activities.’
 Why did we start with electricity markets in Europe? 17

Section of this chapter, topic and relevant regulation Relevant articles

Art. 7(5) and Art. 11(2) state for TSOs and DSOs respectively
that they shall not ‘discriminate between system users or
classes of system users, particularly in favour of [their]
subsidiaries or shareholders.’
However, Directive 96/92/EC left a large margin of choice for Recital 11 of Directive 96/92/EC states that ‘in accordance
the Member States as to how to introduce more competition with the principle of subsidiarity, general principles providing
into their electricity markets. for a framework must be established at Community level, but
their detailed implementation should be left to Member States,
thus allowing each Member State to choose the regime which
corresponds best to its particular situation’; and recital 12
states further that ‘whatever the nature of the prevailing market
organisation, access to the system must be open in accordance
with this Directive and must lead to equivalent economic
results in the States and hence to a directly comparable level of
opening-up of markets and to a directly comparable degree of
access to electricity markets.’
The First Package only required management and accounting Art. 14(3) of Directive 96/92/EC states that ‘Integrated
unbundling. electricity undertakings shall, in their internal accounting,
keep separate accounts for their generation, transmission and
distribution activities, and, where appropriate, consolidated
accounts for other, non-electricity activities, as they would be
required to do if the activities in question were carried out by
separate undertakings, with a view to avoiding discrimination,
cross-subsidization and distortion of competition.’
The Second Package required transmission and distribution Art. 10 and Art. 15 of Directive 2003/54/EC state for TSOs
companies to apply legal unbundling. and DSOs respectively that ‘Where the [TSO, DSO] is part
of a vertically integrated undertaking, it shall be independent
at least in terms of its legal form, organisation and decision
making from other activities not relating to [transmission,
distribution]. These rules shall not create an obligation to
separate the ownership of assets of the [TSO, DSO] from the
vertically integrated undertaking.’
The Third Package required TSOs to be certified by the Art. 9 of Directive 2009/72/EC specifies rules on the
competent NRA under one of three unbundling models. unbundling of transmission systems and TSOs. While Art. 9(1)
implies that the preferred model is ownership unbundling, Art.
9(8) gives Member States the possibility not to apply Art. 9(1)
where ‘on 3 September 2009, the transmission system belongs
to a vertically integrated undertaking’. In such cases, Member
States are given a choice between ownership unbundling and
setting up a system operator or transmission operator which is
independent from supply and generation interests.
Art. 10 of Directive 2009/72/EC lays down the rules for the
designation and certification of transmission system operators.
18 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant regulation Relevant articles

The Third Package also required all the TSOs to create and Art. 4 of Regulation (EC) No 714/2009 requires all TSOs
cooperate via ENTSO-E. to ‘cooperate at Community level through the ENTSO for
Electricity, in order to promote the completion and functioning
of the internal market in electricity and cross-border trade
and to ensure the optimal management, coordinated operation
and sound technical evolution of the European electricity
transmission network.’
Art. 5 of the same Regulation lays down the process of
establishing ENTSO-E.
The selection of TSO and ENTSO-E tasks according to the The tasks of TSOs described are laid out in Art. 7 and Art.
First, Second, Third and Clean Energy Package listed in Figure 8 of Directive 96/92/EC and Art. 9 and Art. 11 of Directive
1.2. 2003/54/EC.
The tasks of ENTSO-E described are laid out in Art. 8 of
Regulation (EC) No 714/2009 and Art. 30 of Regulation (EU)
2019/943.
The Clean Energy Package also requires the establishment of Arts. 52 to 57 of Regulation (EU) 2019/943 lay down
an entity of distribution system operators in the Union (EU provisions for the EU DSO entity.
DSO entity).
The Second Package required Member States to create national Art. 23(1) of Directive 2003/54/EC states that ‘Member
regulatory bodies that are independent of the electricity States shall designate one or more competent bodies with
industry. the function of regulatory authorities. These authorities shall
be wholly independent from the interests of the electricity
industry. They shall … at least be responsible for ensuring
non-discrimination, effective competition and the efficient
functioning of the market …’
The Third Package increased the independence of NRAs from Art. 35 of Directive 2009/72/EC lays out the rules on the
national governments and mandated the establishment of designation and independence of regulatory authorities. Art.
ACER. 35(1) says that ‘Each Member State shall designate a single
national regulatory authority at national level.’ Art. 35(5.a)
specifies that ‘In order to protect the independence of the
regulatory authority, Member States shall in particular
ensure that: (a) the regulatory authority can take autonomous
decisions, independently from any political body, and has
separate annual budget allocations …’
Regulation (EC) No 713/2009 establishes ACER and puts
forward the following motivation in recital 3: ‘it is widely
recognised by the sector … that voluntary cooperation between
national regulatory authorities should now take place within
a Community structure with clear competences and with the
power to adopt individual regulatory decisions in a number of
specific cases.’
The selection of NRA and ACER tasks according to the The tasks of NRAs described are laid out in Art. 23 of
Second, Third and Clean Energy Package listed in Figure 1.3. Directive 2003/54/EC.
The tasks of ACER described are laid out in Chapter II, Arts.
5–11 of Regulation (EC) No 713/2009 and Arts. 3–15 of
Regulation (EU) 2019/942.
 Why did we start with electricity markets in Europe? 19

Section of this chapter, topic and relevant regulation Relevant articles

Network codes and guidelines are adopted in a lengthy Development of network codes: Under the Third Package, the
co-creation process involving the European institutions, network code development process includes multiple steps
ENTSO-E, ACER and many stakeholders from across the as specified in Art. 6 of Regulation (EC) No 714/2009 and
electricity sector. simplified in the following.
Network codes and guidelines carry the same legal weight, First, after having consulted ACER, ENTSO-E and other
are directly applicable, and are subject to the same adoption relevant stakeholders, the European Commission establishes
procedure. an annual priority list for possible network code areas. Second,
Network codes and guidelines differ from each other with at the request of the European Commission, ACER develops
regard to their legal basis (Art. 6 of Regulation (EC) No a non-binding framework guideline setting out principles for
714/2009 for network codes and Art. 8 of the same regulation the development of such network codes and consults with
for guidelines), development process, amendment process and ENTSO-E and other relevant stakeholders. If the European
implementation process. Commission considers that the framework guideline does not
contribute to non-discrimination, effective competition and the
efficient functioning of the market, it may request ACER to
review and resubmit the framework guideline. If ACER fails
to submit a framework guideline, the European Commission
itself elaborates the framework guideline. Third, the European
Commission requests ENTSO-E to develop and submit
to ACER a network code based on the ACER framework
guideline. Fourth, ACER consults relevant stakeholders and
provides ENTSO-E with a reasoned opinion on the network
code. Fifth, ENTSO-E may amend the network code in the light
of the ACER opinion and resubmit it to ACER. Sixth, when
satisfied that the network code is in line with the framework
guideline, ACER submits the network code to the European
Commission and may recommend its adoption. Adoption of
network codes is in the hands of the European Commission.
The European Commission provides reasons in the case where
it does not adopt the network code. Where ENTSO-E has failed
to develop the network code, the European Commission may
request ACER to prepare a draft network code. On its own
initiative or on the failure of ENTSO-E (ACER) to develop
a (draft) network code or on recommendation by ACER, the
European Commission may also adopt one or more network
codes. Where the European Commission proposes to adopt
a network code on its own initiative, it consults ACER,
ENTSO-E and the relevant stakeholders.
Amendment of network codes: Amendments to network codes
under the Third Package may be proposed to ACER by any
person who is likely to have an interest in that network code,
including ENTSO-E, TSOs, system users and consumers.
ACER may also propose amendments on its own initiative.
ACER consults stakeholders and may make a reasoned
proposal for amendment to the European Commission. Taking
account of the ACER proposals, the European Commission
may adopt amendments to any network code.
20 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant regulation Relevant articles

Development and amendment of guidelines: Under the Third
Package, the procedural requirements for guidelines in
accordance with Art. 18 of Regulation (EC) No 714/2009 are
less far-reaching. Art. 18(3) states that, where appropriate,
the Commission can also develop guidelines for areas that are
covered by network codes according to Art. 8(6). Art. 18(3)
states further that the Commission shall consult the Agency
and ENTSO-E for this purpose. When adopting or amending
guidelines, the Commission must, among other things,
ensure that the guideline provides the minimum degree of
harmonization necessary to achieve the goals of the regulation
and does not go beyond what is necessary for this purpose.
Adoption of network codes and guidelines: Under the Third
Package and in accordance with Art. 23 of Regulation (EC)
No 714/2009, network codes and guidelines were adopted
as implementing acts under the ‘old’ comitology procedure,
following the regulatory procedure with scrutiny.
Implementation of network codes and guidelines: As is
described in Box 1.3 of Section 1.1, guidelines include
processes whereby TSOs or NEMOs must develop
so-called ‘Terms and Conditions or Methodologies’ in the
implementation phase.
In theory, network codes and guidelines can cover the same Under the Third Package, Art. 18(6) of Regulation (EC) No
topics. 714/2009 states that ‘Where appropriate, Guidelines providing
the minimum degree of harmonisation required to achieve the
aim of this Regulation shall also specify: … (d) details of the
areas listed in Article 8(6).’Art. 8(6) of the same regulation
lists the areas that network codes shall cover, taking into
account, if appropriate, regional specificities.
Under the Clean Energy Package, in accordance with Art. 61(2)
of Regulation (EU) 2019/943, the ‘Commission is empowered
to adopt guidelines in the areas where such acts could also
be developed under the network code procedure pursuant to
Article 59(1) and (2). Those guidelines shall be adopted in
the form of delegated or implementing acts, depending on the
relevant empowerment provided for in this Regulation.’
Depending on the scope of the respective methodology, the As is set out in Art. 9 of the CACM GL, Art. 4 of the FCA GL,
Third Package foresaw its approval either by all NRAs or the Art. 5 of the EB GL and Art. 6 of the SO GL.
relevant subset of NRAs.
Under the Third Package, in certain cases a decision is to be In accordance with Art. 8(1) of Regulation (EC) No 713/2009,
referred to ACER. in certain cases a decision should be referred to ACER,
namely (a) when the competent NRAs are not able to reach an
agreement within a period of six months from the point in time
when the case was referred to the last of those NRAs or (b) on
a joint request from the competent NRAs.
 Why did we start with electricity markets in Europe? 21

Section of this chapter, topic and relevant regulation Relevant articles

ENTSO-E, ACER and the European Commission have Under the Third Package, relevant provisions, apart from
monitoring, reporting and stakeholder involvement the dedicated articles on network codes and guidelines,
responsibilities related to methodologies. include Regulation (EC) No 714/2009: Art. 8 on the tasks of
ENTSO-E, Art. 9 on monitoring by ACER and Art. 10 on
consultations; and Regulation (EC) No 713/2009: Art. 6 on the
tasks of ACER as regards the cooperation of TSOs, Art. 8 on
the tasks of ACER as regards terms and conditions for access
to and the operational security of cross-border infrastructure,
Art. 10 on consultations and transparency, and Art. 11 on
monitoring and reporting on the electricity and natural gas
sectors.
Under the Clean Energy Package, relevant provisions, apart
from the dedicated articles on network codes and guidelines,
include Regulation (EU) 2019/943: Art. 30 on the tasks of
ENTSO-E, Art. 31 on consultations, Art. 32 on monitoring by
ACER, Art. 41 on transparency, Art. 56 on consultations in the
network code development process, Art. 69 on Commission
reviews and reports; Regulation (EU) 2019/942 Art. 5 on the
tasks of ACER as regards the development and implementation
of network codes and guidelines, Art. 8 on the tasks of ACER
as regards nominated electricity market operators, Art. 24 on
consultations, transparency and procedural safeguards, and Art.
15 on monitoring and reporting on the electricity and natural
gas sectors.
For the first generation of network codes and guidelines, Art. 8(1) of Regulation (EC) No 714/2009 states that ‘The
through ENTSO-E, TSOs were placed in the position of ENTSO for Electricity shall elaborate network codes in the
drafting network codes with regulatory oversight. areas referred to in paragraph 6 of this Article upon a request
addressed to it by the Commission ...’
Following the CEP, the strong role of ENTSO-E in drafting the Art. 59(10) of Regulation (EU) 2019/943 specifies that in the
network codes was reduced. third step of the network code development process described
above, ENTSO-E now drafts the network code based on an
ACER framework guideline guided by a drafting committee
that consists of representatives of ACER, ENTSO-E, where
appropriate the EU DSO entity and NEMOs, and a limited
number of relevant stakeholders.
The CEP also mandates the establishment of an EU DSO entity Art. 52(1) of Regulation (EU) 2019/943 states that
to involve DSOs in the network code and guideline drafting ‘Distribution system operators shall cooperate at Union level
process. through the EU DSO entity, in order to promote the completion
and functioning of the internal market for electricity, and to
promote optimal management and a coordinated operation
of distribution and transmission systems. …’ Arts. 52–57 lay
down rules on distribution system operation, including the
establishment of the EU DSO entity, relevant principal rules
and procedures, and the tasks of the EU DSO entity.
22 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant regulation Relevant articles

Art. 59(3) of the same regulation states that ‘… If the subject
matter of the network code is directly related to the operation
of the distribution system and not primarily relevant to the
transmission system, the Commission may require the EU
DSO entity, in cooperation with the ENTSO for Electricity,
to convene a drafting committee and submit a proposal for
a network code to ACER.’
With the adoption of the CEP, the role of ACER in the In the network code development process following the Third
development phase is expected to increase. Package, as laid down in Art. 6(6–9) of Regulation (EC) No
714/2009 and as described above, ACER is required to provide
a reasoned opinion on the network code submitted to it by
ENTSO-E, after which ENTSO-E may amend the network
code in the light of this opinion and resubmit it to the Agency.
Under the Clean Energy Package, Art. 59(11) of Regulation
(EU) 2019/943 states that, in the fourth step of the network
code development process described above, ACER will
henceforth directly revise and consolidate the network code
and submit the final product to the European Commission:
‘ACER shall revise the proposed network code … and, submit
the revised network code to the Commission within six months
of receipt of the proposal.’ACER must take into account the
views provided by all the parties involved in the drafting of the
proposal and consult the relevant stakeholders on the version to
be submitted to the Commission. Details are also provided in
Art. 5(1.c) of Regulation (EU) 2019/942.
Another change concerns the time interval in which the Under the Third Package, Art. 6(1) of Regulation (EC) No
European Commission is required to compile a priority list for 714/2009 states that ‘The Commission shall, after consulting
new network codes. the Agency, the ENTSO for Electricity and the other relevant
stakeholders, establish an annual priority list identifying the
areas set out in Article 8(6) to be included in the development
of network codes.’
Under the Clean Energy Package, Art. 59(3) of Regulation
(EU) 2019/943 states that ‘The Commission shall, after
consulting ACER, the ENTSO for Electricity, the EU DSO
entity and the other relevant stakeholders, establish a priority
list every three years, identifying the areas … to be included in
the development of network codes.’
ACER now directly decides on the methodologies with Under the Third Package and in accordance with Art. 8(1),
a pan-European scale. decisions are only referred to ACER when (a) the competent
NRAs are not able to reach an agreement within a period of six
months from the point in time when the case was referred to the
last of those NRAs or (b) on a joint request from the competent
NRAs.
 Why did we start with electricity markets in Europe? 23

Section of this chapter, topic and relevant regulation Relevant articles

Under the Clean Energy Package, proposals for common
‘Terms and Conditions or Methodologies’ for the
implementation of network codes and guidelines which require
the approval of all regulatory authorities shall be submitted to
ACER for revision and approval in accordance with Art. 5(2)
of Regulation (EU) 2019/942.
The Clean Energy Package distinguishes between the adoption Under the Third Package and in accordance with Art. 23 of
of network codes and guidelines as implementing or delegated Regulation (EC) No 714/2009, network codes and guidelines
acts. Depending on the type of act, the European institutions were adopted as implementing acts under the ‘old’ comitology
and stakeholders have different rights and possibilities to procedure, following the regulatory procedure with scrutiny.
intervene in the adoption process. The Clean Energy Package distinguishes between the adoption
of network codes and guidelines as implementing or delegated
acts in accordance with Art. 58 of Regulation (EU) 2019/943.
Network Codes as implementing acts: Art. 59(1) of Regulation
(EU) 2019/943 lists the areas in which the Commission is
empowered to adopt network codes as implementing acts.
Adoption of network codes as implementing acts is divided
into two main phases. During the pre-comitology process, the
European Commission undertakes legal and impact assessment
of the implementing act, among other things. During the
comitology phase, the Commission submits draft implementing
acts to a committee composed of Member State representatives.
The committee votes on the draft with three possible outcomes.
In the first case, a qualified majority of Member States votes
in favour of the act resulting in an obligation on the European
Commission to adopt the act. In the second case, a qualified
majority votes against the act, which prohibits the European
Commission from adopting it. In the third case, no qualified
majority for or against exists, which means the European
Commission may adopt the draft.
The CEP empowers the European Commission to adopt
new network codes as implementing acts in the areas
of: (existing, covered by SO GL) network security and
reliability rules; (existing, covered by CACM GL and FCA
GL) capacity-allocation and congestion-management rules;
(existing, covered by EB GL) rules on trading related to the
technical and operational provision of network access services
and system balancing; (new) rules for the non-discriminatory
transparent provision of non-frequency ancillary services; and
(new) rules on demand response, including aggregation, energy
storage and demand curtailment.
24 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant regulation Relevant articles

Network codes as delegated acts: Art. 59(2) of Regulation
(EU) 2019/943 lists the areas in which the Commission
is empowered to adopt network codes as delegated acts.
Adoption as a delegated act can be seen as a fast-track
adoption process, as comitology committees do not exist. The
European Commission prepares and adopts delegated acts after
consulting national expert groups composed of representatives
from each Member State. Citizens and other stakeholders
can provide feedback on the draft of the delegated act. The
European Commission then presents its draft delegated act
simultaneously to both the European Parliament and the
Council without consulting a committee. In the case that the
Parliament and Council do not formulate any objections, the
delegated act enters into force.
The CEP empowers the Commission to adopt new network
codes as delegated acts in the areas of: (existing, covered by
RfG NC, DC NC, HVDC NC) network connection rules;
(existing, covered by ER NC) operational emergency and
restoration procedures in an emergency; (covered partly by
SO GL and Transparency Regulation (EU) No 543/2013) data
exchange, settlement and transparency rules; third-party access
rules (set out in Regulation (EC) No 714/2009 as an area but
not yet covered by a network code); and (new) sector-specific
rules for cyber security aspects of cross-border electricity
flows.
Guidelines: In accordance with Art. 61(2) of Regulation
(EU) 2019/943, the Commission can adopt guidelines in the
network code areas in the form of delegated or implementing
acts. Art. 61 of the same regulation specifies that the adoption
and amendment procedure for guidelines now includes the
obligation for the European Commission to consult ACER,
ENTSO-E, the EU DSO entity and, where relevant, other
stakeholders.
2. Who gets the rights to trade across borders?
Leonardo Meeus with Tim Schittekatte

In this chapter we answer five questions. First, how to deal with historical privileges? Second,
how to implement market-based allocation of transmission rights? Third, how to implement
market coupling in the day-ahead timeframe? Fourth, what about the timeframes before
day-ahead? Fifth, what about the timeframes after day-ahead?

2.1 HOW TO DEAL WITH HISTORICAL PRIVILEGES?

Historically, the rights to trade across borders were granted to utilities, often state-owned
vertically integrated utilities. The First Directive 96/92/EC already stated that the newly
established transmission system operators (TSOs) had to provide different network users with
non-discriminatory access to their networks. Despite this provision, the historical privileges
of the utilities were maintained. They had long-term contracts with neighbouring utilities that
included transmission rights. These contracts, which pre-dated the market integration process,
had not yet expired and were kept in place.
This led to a landmark court case. In that case, VEMW, the organization representing the
interests of large energy consumers in the Netherlands, the Amsterdam Power Exchange
(APX) and ENECO, a large Dutch utility, challenged the decision of DTE, the Dutch regu-
lator, to reserve a significant proportion of the rights to trade across the border for SEP, the
former national vertically integrated utility. SEP used these so-called transmission rights to
execute its existing long-term contracts with utilities across the border. In 2000, 1500 MW of
the available 3200 MW of the rights were reserved for SEP contracts, which would reduce to
900 MW in 2001 and to 750 MW from 2005 to 2009. The European Court of Justice (ECJ)
decided that this undermined potential access to the market by new players and protected the
position of the incumbent. The Dutch regulator’s decision was found to be incompatible with
the First Directive and so annulled. Although the First Directive allowed Member States to
request a transitional exemption from the relevant article in the legislation, this had not been
done by the Netherlands. Other national regulatory authorities (NRAs) used the ECJ decision
to take steps to remove transmission right privileges.1
The next challenge was to allocate the freed-up transmission rights in a non-discriminatory
way. Figure 2.1 gives an overview of the transmission right allocation methods applied in the
EU in 2004. Priority lists were said to give priority to whoever made it onto the list, often
incumbent utilities and large industrial consumers. Pro-rata rationing implied that whoever
asked for more got more. Explicit auctions and market splitting were the first market-based
approaches to allocate transmission rights. They will be discussed in more detail in the next

25
26 Evolution of electricity markets in Europe

Note: The acronyms used stand for the following. MO: Morocco (non-EU); P: Portugal; E: Spain; I: Italy; F:
France; CH: Switzerland (non-EU); A: Austria; SL: Slovenia; G: Greece; H: Hungary; SK: Slovakia; CZ: Czech
Republic; Pl: Poland; D: Germany; L: Luxembourg; R: Russia (non-EU); DK (E): East Denmark; DK (W): West
Denmark; FI: Finland; S: Sweden; N: Norway (non-EU); B: Belgium; NL: Netherlands; UK: United Kingdom; IR:
Ireland.
Source: Based on European Transmission System Operators (ETSO) (2004).

Figure 2.1 Implementation of different allocation methods for cross-border transmission

rights in Europe in 2004

sections. Finally, please note that on several borders there is not a unique capacity allocation
method or congestion management mechanism jointly applied by the two TSOs involved.

2.2 HOW TO IMPLEMENT MARKET-BASED ALLOCATION OF

TRANSMISSION RIGHTS?

Regulation (EC) No 1228/2003 was included in the Second Package and required a market-based
approach to the allocation of transmission rights. In what follows we describe the evolution
from explicit auctions to implicit auctions – or market coupling – in Europe.
First, explicit auctions. Under this approach, TSOs auction transmission rights to the highest
bidders separately from the trading of energy. Several auctions are held, from the year-ahead
to the day-ahead timeframe. A few years after the adoption of Regulation (EC) No 1228/2003,
explicit auctions became the dominant model for allocating transmission rights in Europe. Due
to the separation of the auctions of transmission rights and energy trading, coordination issues
arose. More precisely, to be able to bid for transmission rights, traders had to predict hourly
price differences in different countries, which turned out to be very difficult. The 2007 Sector
Inquiry estimated that the lost opportunities to trade across the German–Dutch border in 2004
were as high as €50 million, or half the total value.2 Figure 2.2 shows the details. Each dot
 Who gets the rights to trade across borders? 27

Source: European Commission (2007).

Figure 2.2 Explicit cross-zonal allocation: hourly price difference between Germany
and the Netherlands (x-axis) versus the hourly sum of nominated net flows
from Germany to the Netherlands in 2004 (y-axis)

represents the situation in the market at a certain hour. Two main problems can be observed.
First, quadrant 2 and quadrant 4 show hours in which traders moved energy across the border
earning the price spread, but they often did not capture all the opportunities. There are many
hours in which not all the transmission rights were used although there was still a positive
price spread. Second, quadrants 1 and 3 show hours in which traders paid for transmission
rights and then moved energy across the border in the wrong direction, losing money. In
this example, for about 40 per cent of the hours the electricity price in Germany was higher
than in the Netherlands but the electricity was flowing towards the Netherlands (quadrant 1).
The opposite also happened, with electricity flowing towards Germany when the price in the
Netherlands was higher (quadrant 3).
Second, there was an evolution towards implicit auctions, or market coupling. The solution
that emerged was to give the transmission rights to power exchanges. Instead of allocating
them in a separate auction to cross-border traders, they were integrated into the clearing of the
day-ahead auction organized by power exchanges. The name that was initially given to this
solution was ‘implicit auctions’, but it then changed to ‘market coupling’ because the solution
implied that the power exchange day-auctions became coupled, as we will discuss further in
the next section.
28 Evolution of electricity markets in Europe

2.3 HOW TO IMPLEMENT MARKET COUPLING IN THE

DAY-AHEAD TIMEFRAME?

This section is divided in two subsections. We first describe the evolution from regional ini-
tiatives to so-called single day-ahead coupling (SDAC). We then discuss the issues that this
solution left open. Power exchanges are key actors with regard to market coupling; a descrip-
tion of the evolution of power exchanges in Europe can be found in Annex 2A.1.3

2.3.1 From Regional Initiatives to Single Day-ahead Coupling

In this subsection, we first discuss three regional initiatives that led to SDAC: market splitting,
trilateral market coupling and volume coupling. As Table 2.1 summarizes, we discuss the
number of power exchanges that were involved in the initiative, whether they share their order
books and by whom the optimization algorithm is run. We conclude with a description of the
current status of SDAC implementation.
First, market splitting. Nord Pool’s approach was called market splitting because the
algorithm worked in two steps. In the first step, the Nordic system price was calculated. The
system price is the price that would apply in the whole region if the resulting cross-border
flows were feasible. If there were not enough cross-border transmission capacity available to
accommodate all trades, the second step was to split the Nordic market into smaller markets
with different prices, hence ‘market splitting’. It all started with Nordic market splitting. In
this approach, there was one power exchange – Nord Pool – with one order book and one
optimization algorithm to calculate prices for the whole Nordic region.
Second, trilateral market coupling. The starting point for this project was three power
exchanges (APX, Belpex and Powernext) and three TSOs (TenneT, Elia and RTE) which
wanted to implement market coupling without consolidating the exchanges into one exchange.
Instead of sharing all the information in their order books, they wanted to run an optimization
algorithm based on net export curves, which was a way to aggregate their order book infor-
mation. The algorithm would then decide the trade volumes and directions across the borders,
and once these were fixed the exchanges would continue to calculate their own set of prices for
their order books. It was a very elegant idea, and we did research to help develop the concept,
but our research also showed that the approach had its limitations.4 We were therefore not
surprised that after a piloting phase with detailed simulations the project partners decided to
abandon the approach and went for a slightly more centralized version of market coupling.
This was approved by the regulators and went live in 2006. The exchanges did not consolidate
into one exchange, but they did agree to share their full order books and to let one optimization
algorithm calculate the prices. The compromise was that they would take turns to run the algo-
rithm or they would run it in parallel to build some resilience into the procedure.
Third, volume coupling. Two years after the approach based on net export curves in the
trilateral market project was abandoned, it resurfaced in an initiative between Nord Pool (East
Denmark) and EEX (Germany). Their approach was called volume coupling, as opposed to
price coupling. It was short-lived because the problems that had been anticipated in trilateral
market coupling and were analysed in our research unfortunately were realized. After several
attempts to fix it, the project was stopped. For a full account of what happened, with an
analysis of the performance of volume coupling in each step of the way, see our research on
 Who gets the rights to trade across borders? 29

Table 2.1 Different design choices for the implementation of market coupling

Market Trilateral market Volume coupling Single day-ahead coupling

splitting coupling
Number of power One (Nord Three (Belpex, APX Two Many, now certified as NEMOs
exchanges involved Pool) and Powernext) (EEX and Nord
Pool)
Order books shared Only one Yes No Yes
order book
Optimization Nord Pool Rotational – To validate Both One NEMO runs the algorithm on
algorithm run by the others do parallel a rotational basis. The other NEMOs
runs have the right to do parallel runs to
validate the results

the topic.5 This short-lived experiment settled the competition between different market cou-
pling implementations in favour of trilateral market coupling, which developed into SDAC.
A simple numerical example showing how prices are set and cross-border capacity is allocated
between two market-coupled countries can be found in Annex 2A.2.
Fourth, the status of SDAC. The Capacity Allocation and Congestion Management
Guideline (CACM GL), which was adopted in 2015, made day-ahead market coupling
binding for all. Trilateral market coupling has been growing, and at the time of writing 22
countries representing close to 90 per cent of European electricity consumption are already
coupled. More are expected to join in 2020: Greece, the Czech Republic, Slovakia, Hungary
and Romania.6 The optimization algorithm that is used is called the Pan-European Hybrid
Electricity Market Integration Algorithm (EUPHEMIA), and it originated in the trilateral
market coupling project. The operation of the algorithm is called the market coupling operator
(MCO) function in the CACM GL. The MCO function is jointly operated by all the participat-
ing power exchanges. To be able to participate, exchanges have to be certified as Nominated
Electricity Market Operators (NEMOs).

2.3.2 Open Issues

In this subsection, we discuss three open issues: the governance of NEMOs and the MCO
function, cost sharing and the functioning of the optimization algorithm EUPHEMIA.
The first open issue is governance. The default option described in the CACM GL is to have
competing power exchanges in each Member State. However, if a national legal monopoly for
day-ahead and intraday trading services existed in a Member State at the time of entry into
force of the CACM GL, the Member State could decide to continue with a monopolistic power
exchange. At the time of writing, Greece, Ireland, Italy, Spain, Portugal, Hungary, Bulgaria,
Slovakia, the Czech Republic and Romania have designated one monopolistic NEMO and all
the other Member States allow multiple NEMOs to compete in their markets. Competitive
NEMOs designated in one Member State also have the right to offer trading services with
delivery in another Member State where a competitive model is implemented, unless an excep-
tion is justified. Recently, multi-NEMO arrangements (MNAs) have been set up, enabling
market coupling with more than one power exchange per country.
30 Evolution of electricity markets in Europe

From the operational perspective, currently eight NEMOs are running the MCO function on
a rotational basis. Each day one NEMO runs the system (acting as the ‘operator’), one acts as
the formal ‘coordinator’ (e.g. announcing official results and calling the incident committee
when issues arise) and another acts as ‘hot backup’. The remaining NEMOs have the right
to compute the same results in parallel. This has put power exchanges in a position in which
they must collaborate while they are also competing. The governance of NEMOs and the
MCO function was on the agenda of the Florence Forum in 2018. The European Network of
Transmission System Operators for Electricity (ENTSO-E) called for a stronger separation of
the MCO function and the competitive business activities of NEMOs, preferably with more
TSO involvement in the control of the MCO function. The Agency for the Cooperation of
Energy Regulators (ACER) acknowledged the issues and hinted at the possibility of having
a single independent MCO entity. The solution suggested by ACER was included in the
CACM GL as a possible option if the current option does not work well. Finally, in a report
published just after the Florence Forum, the Commission decided that it was too early for
action.7
The governance issue has a history. During the development of the CACM GL, the question
arose of whether power exchanges need to be regulated as monopolies now that they have
a monopoly on cross-border trade in the day-ahead timeframe. Our research concluded that the
competitive model for market infrastructure has its merits. Some of the activities of a power
exchange, such as the provision of the interface between traders and the market and all sorts
of settlement arrangements, are not necessarily monopolistic activities. Competing means they
can differentiate and innovate in their services and also in their membership fees and trade
commissions. We also wrote that it is necessary to avoid market coupling becoming a cartel
of power exchanges, which did not make us very popular at the time. In 2014, the European
Commission imposed fines of about €6 million on the two leading European power exchanges,
EPEX SPOT and Nord Pool. They had agreed not to compete in the spot market (day-ahead
and intraday) and to divide the European territory between them. In other words, the govern-
ance of NEMOs and the MCO function is an important issue to continue to monitor.8
The second open issue is cost sharing. The costs of operating and developing the MCO
function are shared by the NEMOs. The CACM GL states that TSOs may contribute to the
MCO-function-related costs of the NEMOs concerned but they are not obliged to. In the case
that a TSO does not contribute to the costs or does not cover all the costs, the NEMOs are enti-
tled to recover residual MCO-function costs by means of regulated fees or other appropriate
mechanisms unless the costs are unreasonable. In short, the sharing keys between NEMOs and
TSOs for MCO-function-related costs are national, while NEMOs can cover multiple Member
States. As not necessarily the same NEMOs are competing in each Member State, due to the
non-harmonized national sharing keys some NEMOs might need to top up their fees more than
others because of the costs of MCO-related activities.
The third open issue is the algorithm. Technical issues came high on the agenda after the
first major incident on 7 June 2019. Due to a software bug at EPEX SPOT – the biggest
power exchange in central Europe, operating in countries such as Austria, Belgium, France,
Germany and the Netherlands – its order books were not included in the market coupling algo-
rithm EUPHEMIA. This meant that trade could not be scheduled across the borders of these
countries. The fall-back solution was to organize local markets without cross-border trade. As
a result, prices were unusually low in some countries and extremely high in others.9
 Who gets the rights to trade across borders? 31

At the time of writing, EUPHEMIA only has ten minutes to perform its market coupling
calculations. Even though the incident on 7 June 2019 was said to have been caused by a soft-
ware bug at EPEX SPOT, there are also technical issues with the algorithm itself. These issues
are attracting more attention because the single-market-coupling approach in Europe relies on
this one algorithm. As more countries have been added to the market coupling initiative, insuf-
ficient efforts have been made to reduce the complexity of the market that has to be handled
by the algorithm. The CACM GL required NEMOs to submit a joint proposal for the bidding
formats they will continue to use. In their proposal, which was accepted, they simply listed
all the bidding formats that are currently used. These are a mix of multi-part orders with bids
that correspond to the technical constraints of thermal power plants, including start-up costs
and ramping constraints, and block orders, which allow traders to make their bids indivisible
and to link them across periods. Multi-part orders are typical of power-pool type markets,
while block orders are typical of power exchanges. EUPHEMIA currently has to deal with
both complexities. In power-pool markets, complex bidding formats are typically combined
with complex pricing, which means that side payments are used to balance supply and demand
in the market on top of the clearing price. This implies that some market players get an addi-
tional side payment to avoid losing money, while the others pay or receive the clearing price
corrected with an uplift to recover the money paid in side payments. Power exchanges do not
do side payments; they instead allow their markets to reject block bids that are in-the-money
at the prevailing clearing price (so-called paradoxically rejected blocks). Side payments would
require a change in the CACM GL, but they would simplify the algorithm. Our research and
that of colleagues has indeed demonstrated that complex pricing can reduce the complexity of
the algorithm, but we have come to different conclusions on whether it should be done or not.
Our colleagues argued in favour of complex pricing while we argued that the gains in terms
of trade would be relatively small in comparison to the side payments that would need to be
administered.10
Stakeholders are under pressure to come up with solutions because the complexity is only
expected to increase. More countries will be added to SDAC, and the Clean Energy Package
includes provisions that will increase the granularity of day-ahead markets from hourly to
half-hourly or even 15 minutes.
Note finally that more progress has been made on another technical issue: the minimum and
maximum clearing prices that are used by the different power exchanges have been harmo-
nized. Following the CACM GL, all NEMOs were first asked to come up with a joint proposal
which the NRAs had to unanimously approve. The NRAs could not reach an agreement and
therefore requested ACER to adopt a decision. ACER decided that the harmonized maximum
price in the day-ahead market should be €3000/MWh and the minimum price −€500/MWh.
ACER avoided the maximum clearing price being able to act as a price cap by making it
dynamic. In the event that the clearing price exceeds a value of 60 per cent of the previously
set maximum clearing price, the maximum clearing price is increased by €1000/MWh.

2.4 WHAT ABOUT THE TIMEFRAMES BEFORE DAY-AHEAD?

The timeframes before day-ahead are referred to as forward and futures markets. These mainly
involve bilateral deals or over-the-counter (OTC) trading. Power exchanges also play a role,
but not by organizing auctions as they do in the day-ahead timeframe. Instead, they offer
32 Evolution of electricity markets in Europe

platforms that enable continuous trade in standardized future contracts, for example one-year
and three-year contracts. This means that there are no power exchange auctions that can be
coupled in the timeframes before day-ahead. Cross-border long-term transmission rights can
only be allocated in explicit auctions, the market-based solution that was abandoned in the
day-ahead timeframe. Indeed, TSOs organize explicit auctions for at least monthly and yearly
transmission right contracts. They have also started to collaborate via the Joint Allocation
Office (JAO). JAO is a joint service company of currently 20 TSOs from 17 countries with
harmonized auction rules and timings, which helps traders reduce their transaction costs in
procuring transmission rights. With the Forward Capacity Allocation Guideline (FCA GL),
which entered into force in 2016, this bottom-up voluntary initiative became the European
platform. Following the FCA GL, in 2017 all the NRAs approved a methodology that pro-
posed JAO becoming the single allocation platform for the whole of Europe. In what follows,
we discuss the two main open issues regarding long-term transmission rights: the length and
the types of contracts.
The first open issue is the length of contracts. To split the available transmission rights over
the different long-term timeframes and contract lengths, the FCA GL foresees the need to
develop methodologies at the regional level. At the time of writing, these methodologies are
being discussed and have not yet been approved. Traders have made it clear that they prefer to
have capacity offered year-ahead or more. Most TSOs propose a somewhat gradual offering
by dividing the capacity over the different timeframes. The incentives for TSOs depend on the
compensation they have to pay to market parties if they have to curtail long-term transmission
rights because they face problems in their networks, and on whether NRAs allow TSOs to
recover these payments through their grid tariffs. No compensation payments is not an option
because then market parties would have a hedge that they could not rely on because it can be
curtailed without any consequences for the TSO that issued the hedging product. This has been
an issue in the past and has been addressed in the FCA GL. In the FCA GL, two causes for the
curtailment of long-term transmission rights are distinguished. If what happens is considered
force majeure, the price of the right in the original auction is refunded. The determination of
whether an event classifies as force majeure is, however, still done at the national level. More
precisely, the national regulatory authority of the TSO invoking a force majeure event has to
assess whether the event qualifies as force majeure. If it is not force majeure, that is, the TSO
curtails long-term transmission rights to ensure that all flows remain within the operational
security limits, the compensation is the lost opportunity, which is the day-ahead price spread.
In this case, the TSOs concerned might propose introducing a cap on the total compensation,
which is further specified in the FCA GL.
The second open issue is the type of contract. Most TSOs started by auctioning so-called
physical transmission rights (PTRs). A trader that buys such a right can trade across the border
and nominate that trade to the TSO, which then subtracts this capacity from the overall volume
of transmission rights that remain for the other timeframes. If the trader decides not to use the
right, it is compensated for the value of the right in a day-ahead auction, where other traders
might be willing to pay for it (use-it-or-sell-it, UIOSI). If the day-ahead stage applies market
coupling, the price difference across the border is the implicit price for the transmission right.
Most TSOs have already converted or are converting to another type of long-term transmission
right referred to as a financial transmission right (FTR). With FTRs, use-it-or-sell-it becomes
sell-it-without-the-possibility-of-using-it. Traders still hedge against the day-ahead price
 Who gets the rights to trade across borders? 33

differences between countries, but they cannot nominate a cross-border flow ahead of the
day-ahead timeframe. Hence the name financial, because the physical element is no longer
there. At the time of writing, FTRs are in place on nine borders in the EU and implementation
of FTRs is planned on an additional eight borders.11
The FCA GL leaves it open whether PTRs or FTRs are used. However, among other things,
it is required that marginal pricing is applied in the auctions, that for both PTRs and FTRs
harmonized allocation rules are followed, and that the two types of transmission right cannot
be applied in parallel on one border. Regulators can adopt a coordinated decision not to issue
PTRs or FTRs on a border when they can show that there is no need for hedging or by ensuring
that there are other cross-border hedging instruments available in the market. For example, in
the Nordics an instrument called electricity price area differentials (EPADs) is in place. EPAD
contracts hedge the difference between the price in a certain location and the Nordic system
price. The Nordic system price is the price that would have emerged if there were no conges-
tion in the Nordic system. This system price is a legacy from the market splitting approach and
does not represent an actual location in the Nordic system. Italy does something similar with
an instrument that hedges the difference between the price in a certain location for supply and
the ‘unique’ national price for demand, that is, Prezzo Unico Nazionale (PUN).

2.5 WHAT ABOUT THE TIMEFRAMES AFTER DAY-AHEAD?

The timeframes after day-ahead are intraday and balancing markets. Power exchanges play
a role in the intraday stage by organizing continuous trading platforms, and in some cases also
auctions. In the day-ahead and forward markets the debate was mainly over how to allocate
transmission rights, while for the intraday stage the debate was over keeping the borders open
long enough for intraday trade to become international. The CACM GL prescribes that the
intraday cross-border gate closure shall be at most one hour before delivery. After the intraday
cross-border gate closure, national intraday markets often remain open until even closer to real
time.
Progress in the intraday stage has been slower than in the day-ahead stage. The volumes that
are traded are smaller so less money can be made by organizing and participating in intraday
trade, but intraday markets are important because they allow market parties to avoid imbal-
ances. This is especially important for new players and renewable energy technologies, which
would otherwise be exposed to high balancing costs. We will come back to this interaction in
Chapter 5 on balancing markets. In what follows, we focus on the two intertwined open issues
related to intraday markets: the transmission right allocation method and reservation.
Continuous trade became the dominant model in Europe for intraday in a context where
intraday was less important and trade volumes were so low that auctions were not considered
a feasible option. Now, this intraday continuous trade can also be cross-border using transmis-
sion rights. A main open issue is how to allocate these transmission rights. For most borders,
the current practice is that intraday transmission rights are used free of charge by whoever is
matched first on the continuous trade platform until the rights are no longer available or the
border is closed (i.e. one hour before delivery). After that, the matching of traders can continue
locally. However, first-come-first-served is not a market-based allocation method. Therefore,
no transmission rights are reserved for the intraday stage; only the rights that have not been
used in the day-ahead stage are allocated. Another option is to organize explicit auctions for
34 Evolution of electricity markets in Europe

intraday transmission rights complementing continuous trade, which at the time of writing is
done on several borders. However, we know from experience in the day-ahead market that
explicit allocation is not without flaws.
The CACM GL pushes continuous trade in the intraday timeframe, and auctions are tolerated
as complementary regional arrangements. The cross-border intraday market project (XBID),
based on continuous trading with first-come-first-served allocation of transmission rights, has
been formalized as the single intraday coupling (SIDC) to be applied by all Member States.
At the time of writing, XBID is composed of members from 21 European countries (‘the first
and second waves’). However, at the same time, the CACM GL requires a single methodology
for intraday cross-zonal pricing reflecting market congestion through an implicit allocation
method. Explicit allocation is only allowed as a transitional complementary arrangement. The
implementation of this methodology has been challenging. The NRAs could not agree on how
to do this, so ACER had to decide. Finally, in 2019 ACER decided that three pan-European
auctions will be introduced on top of continuous trading in the intraday timeframe. As soon
as there are auctions in the intraday timeframe, transmission rights can be allocated efficiently
through market coupling and will no longer be allocated for free. The debate on reservation
of transmission rights for this timeframe can even be reopened. How this will play out is very
much an open issue.

2.6 CONCLUSION

In this second chapter, on who gets the rights to trade across borders, we have answered five
questions.
First, how to deal with historical privileges? In Europe, they had remained in place for
a relatively long period until they were challenged in court. Finally, the European Court of
Justice found these privileges were incompatible with the First Electricity Directive. This
landmark court case opened the door for other countries to also abolish them. The accepted
idea was to introduce market-based allocation of transmission rights. However, we had to
wait for Regulation (EC) No 1228/2003, adopted as part of the Second Energy Package, for
market-based allocation of transmission rights to be made mandatory on all EU borders.
Second, how to implement market-based allocation of transmission rights? Regulation (EC)
1128/2003 did not specify which market-based approach should be used. In Europe, many dif-
ferent approaches were tried before converging towards market coupling. Most borders started
with explicit auctions for transmission rights as this solution did not require many changes to
national electricity markets. However, it was quite quickly shown that explicit auctions had
strong deficiencies in the day-ahead timeframe.
Third, how to implement market coupling in the day-ahead timeframe? Different imple-
mentations emerged through regional initiatives. The EU network codes and guidelines made
the implementation of market coupling legally binding through the CACM GL. At the time
of writing, day-ahead market coupling covers 22 European countries, representing more than
90 per cent of EU electricity consumption. The governance of the market coupling operator
(MCO) function and the performance of the algorithm are the main open issues today.
Fourth, what about the timeframe before day-ahead? The network codes and guidelines,
more specifically the FCA GL, also impacted cross-border long-term transmission rights.
Currently, TSOs are obliged to issue transmission rights at least month-ahead and year-ahead
 Who gets the rights to trade across borders? 35

on a joint allocation platform. However, there are still ongoing discussions on how to divide
long-term transmission rights between the year-ahead and month-ahead auctions. Traders
want to have as many rights as possible allocated far ahead of delivery, while TSOs generally
propose dividing rights more equally over the timeframes. Moreover, the network codes and
guidelines do not settle which kind of transmission right should be issued. Historically, on
most borders physical transmission rights were in place. Currently, we are seeing a transition
from physical to financial transmission rights in Europe.
Fifth, what about the timeframes after day-ahead? Historically, due to low liquidity,
intraday markets were organized as continuous trading platforms with few options to trade
cross-border. Important progress has been made with the ongoing implementation of XBID,
the single intraday market coupling solution. ACER has decided to complement XBID with
three pan-European intraday auctions. Currently, no cross-border capacity is planned to be
reserved for intraday.

NOTES
1. More in-depth discussion of this case can be found in Hancher (2006).
2. The competition authority of the European Commission opened a Sector Inquiry into the func-
tioning of the European energy markets after significant price increases in the European electricity
wholesale markets (European Commission 2007). Generally, the idea was that by integrating
markets competition can be fostered. In this regard, Gilbert et al. (2004) show that the way that
transmission rights are allocated to generators can also have a strong impact on whether market
power can be mitigated or not.
3. Annex 2A.1 is based on our own work and that of our colleagues. Boisseleau (2004) was one of
the pioneers discussing the importance of power exchanges. In Meeus et al. (2005b) we discuss the
early development of the EU electricity market and the role of power exchanges. In Meeus (2011b)
we focus on the governance and business models of power exchanges and power pools in Europe.
For market statistics, refer to ACER and CEER (2019) and DG Energy (2019).
4. EuroPex (2003) introduced the concept of net export curves. In Meeus et al. (2005a) we argue that
the concept can work if we only consider simple and single-period orders. But, as discussed in
Meeus (2006), the problem is block orders, which would require many iterations and the quality of
the solution would suffer.
5. In Meeus (2011a) we focus on the experience with volume coupling. We show that volume cou-
pling initially performed worse than the situation without coupling. The implementation of volume
coupling was then changed, which slightly improved the performance but not enough to save the
project. We also explain mathematically why what happened was to be expected.
6. More specifically, two coupling projects are in parallel operation, namely the Multi-Regional
Coupling (MRC) and the 4M Market Coupling (4MMC) projects. At the end of 2019, the MRC
connected 22 countries. Furthermore, Greece is also expected to be coupled through the Greece–
Italy interconnector in 2020. The other coupling project, the 4MMC, covers the Czech Republic,
Slovakia, Hungary and Romania. Both projects are expected to be merged in 2020 (ENTSO-E
2019a).
7. In a report published just after the Florence Forum in 2018, the European Commission (2018) wrote
that ‘the Commission sees a need to continue the discussion on the challenges faced so far and
assess the various options for a potential change in the governance of the MCO function.’ All the
presentations at the Florence Forums can be accessed online through the European Commission’s
dedicated website.
8. In Meeus (2011b) we distinguish between merchant and cost-of-service-regulated power exchanges.
We show that these models each have pros and cons by referring to experiences in financial markets.
We also show that more responsibility for power exchanges comes with market coupling, which
might require a governance model to be put in place. At the time of writing we have two main
36 Evolution of electricity markets in Europe

worries. First, that power exchanges may not implement market coupling properly to protect their
own business, as we found in Meeus (2011a). Second, asking them to collaborate to organize market
coupling should not result in a cartel. The press release on the antitrust case of power exchanges can
be found in European Commission (2014).
9. For more information, consult the NEMO Committee (2019) report on the incident.
10. In Meeus et al. (2009) we use simulations based on market data from APX to compare the results of
an algorithm with and without side payments. Madani et al. (2018) further develop the algorithm,
which can be used for market coupling with complex pricing, arguing that it would be a simpler way
of clearing markets.
11. This is reported in ENTSO-E (2019b). ACER (2019b) keeps track of the status of long-term trans-
mission rights on its website.

REFERENCES
ACER (2017a), ‘Decision No 04/2017 of the Agency for the Cooperation of Energy Regulators of 14
November 2017 on the NEMOs’ Proposal for Harmonised Maximum and Minimum Clearing Prices
for the SDAC’.
ACER (2017b), ‘Decision No 05/2017 of the Agency for the Cooperation of Energy Regulators of 14
November 2017 on the NEMOs’ Proposal for Harmonised Maximum and Minimum Clearing Prices
for the SIDC’.
ACER (2019a), ‘Decision No 01/2019 of the Agency for the Cooperation of Energy Regulators of 24
January 2019 on Establishing a Single Methodology for Pricing Intraday Cross-Zonal Capacity’.
ACER (2019b), ‘Cross-zonal hedging status’, accessed at https://​www​.acer​.europa​.eu/​en/​Electricity/​
MARKET​-CODES/​FORWARD​-CAPACITY​-ALLOCATION/​Pub​_Docs/​Crosszonal hedging
status.pdf.
ACER and CEER (2019), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2018 – Electricity Wholesale Markets Volume’.
All NEMOs (2017a), ‘All NEMOs’ Proposal for Products that Can Be Taken into Account by NEMOs
in Intraday Coupling Process in Accordance with Article 53 of the CACM GL’, published on 13
November 2017.
All NEMOs (2017b), ‘All NEMOs’ Proposal for Products that Can Be Taken into Account by NEMOs
in Single Day-Ahead Process in Accordance with Article 40 of the CACM GL’, published on 13
November 2017.
All NEMOs (2017c), ‘All NEMOs’ Proposal for the MCO Plan’, published on 13 April 2017.
Boisseleau, F. (2004), ‘The Role of Power Exchanges for the Creation of a Single European Electricity
Market: Market Design and Market Regulation’, PhD Thesis, Delft University of Technology/Paris
IX Dauphine University.
DG Energy (2019), ‘Quarterly Report on European Electricity Markets’, Market Observatory for Energy
Q2/2019, vol. 12.
ENTSO-E (2019a), ‘Market Report 2019’.
ENTSO-E (2019b), ‘Type of LTTRs – current status’, presentation at the European Market Stakeholder
Committee on 02/07/2019, accessed at https://​docstore​.entsoe​.eu/​Documents/​Network codes doc-
uments/Implementation/stakeholder_committees/MESC/2019-07-02/4.1_190702_MESC_status_
LTTRs.pdf.
ETSO (2004), ‘An Overview of Current Cross-Border Congestion Management Methods in Europe’.
European Commission (2007), ‘DG COMP Report on Energy Sector Inquiry’.
European Commission (2014), ‘Antitrust: Commission fines two power exchanges € 5.9 million in cartel
settlement’, press release of 5 March 2014.
European Commission (2018), ‘Report from the European Commission to the Council and the European
Parliament on the Development of Single Day-Ahead and Intraday Coupling in the Member States and
the Development of Competition between NEMOs in Accordance with Article 5(3) of Commission
Regulation 2015/1222 (CACM)’.
EuroPex (2003), ‘Using implicit auctions by power exchanges to manage cross exchanges: decentralized
market coupling’, presentation by B. Den Ouden, President of EuroPex at the Florence Forum.
 Who gets the rights to trade across borders? 37

Gilbert, R., K. Neuhoff and D. Newbery (2004), ‘Allocating transmission to mitigate market power in
electricity networks’, RAND Journal of Economics, 35 (4), 691–709.
Hancher, L. (2006), ‘Case C-17/03, VEMW, APX en Eneco Nv v. DTE’, Common Market Law Review,
43 (4), 1125–44.
Madani, M., C. Ruiz, S. Siddiqui and M. Van Vyve (2018), ‘Convex hull, IP and European electricity
pricing in a European power exchanges setting with efficient computation of convex hull prices’,
ArXiv Preprint ArXiv:​1804​.00048, accessed at https://​arxiv​.org/​abs/​1804​.00048.
Meeus, L. (2006), ‘Power Exchange Auction Trading Platform Design’, PhD Thesis, KU Leuven.
Meeus, L. (2011a), ‘Implicit auctioning on the Kontek Cable: Third time lucky?’, Energy Economics, 3
(33), 413–18.
Meeus, L. (2011b), ‘Why (and how) to regulate power exchanges in the EU market integration context?’,
Energy Policy, 39 (3), 1470–5.
Meeus, L., T. Meersseman, K. Purchala and R. Belmans (2005a), ‘Implicit auctioning of network capac-
ity by power exchanges using net export curves’, EEM Conference Paper.
Meeus, L., K. Purchala and R. Belmans (2005b), ‘Development of the internal electricity market in
Europe’, Electricity Journal, 18 (6), 25–35.
Meeus, L., K. Verhaegen and R. Belmans (2009), ‘Block order restrictions in combinatorial electric
energy auctions’, European Journal of Operational Research, 196 (3), 1202–6.
NEMO Commitee (2019), ‘SDAC Report on the “Partial Decoupling” Incident of June 7th 2019’,
Version: 1.0, accessed at http://​www​.nemo​-committee​.eu/​assets/​files/​sdac​-report​-on​-decoupling​.pdf.
38 Evolution of electricity markets in Europe

2A.1 ANNEX: THE EVOLUTION OF THE ROLE OF POWER

EXCHANGES IN EUROPE

In this annex we describe how the role of power exchanges evolved in Europe. First, we
introduce the slow start of power exchanges. Second, we explain the increased focus on
them. Third, we discuss the circumstances that led to a consolidation of the number of power
exchanges across Europe.
First, the slow start of power exchanges. The evolution of power exchanges in Europe
started in 1993 with Statnett Marked AS in Norway. Three years later Sweden joined the
initiative, which was renamed as Nord Pool ASA. After, Nord Pool extended to Finland
and Denmark in 1998 and 2000 respectively. In Spain, OMEL was founded in 1998. In
the Netherlands, the Amsterdam Power Exchange (APX) was launched in 1999. In 2000,
Germany saw its first power exchange (APX Deutschland, APXDE) launch and then cease
operation after only a couple of months without any trading. In the same year, the Leipzig
Power Exchange (LPX) and later the European Energy Exchange (EEX) were created. The
French exchange Powernext was launched in 2001, and many other countries followed
after. Power exchanges were market infrastructure set up by market parties, financial market
institutions, TSOs or a combination of private actors. Over-the-counter (OTC) markets and
organized exchanges complement each other but also compete for trading volume, which helps
to reduce transaction costs for traders. Power exchanges are trading platforms that facilitate
anonymous trade between market parties. By acting as a counterparty in all transactions and
clearing all trades either themselves or through their clearing houses, they greatly reduce the
counterparty risk for market participants. Power exchanges also enhance market transparency
as prices and volumes are published through the platform, while details of OTC trades remain
with the negotiating parties. The closer to real time, the more specific the needs of the market
participant are and the more difficult it is to find the right counterparty. Despite the early
movers mentioned above, it took a long time for all the markets to have a power exchange up
and running. Still today, OTC trade constitutes the bulk of electricity trade.
Second, the increased focus on power exchanges. Having a healthy and well-functioning
exchange became a benchmark in national market functioning. The Sector Inquiry of 2007
used several indicators to measure the performance of power exchanges, such as the number of
players, traded volumes, the price-setting frequency of certain generators and price volatility.
Later, price resilience was also added as an indicator. Not surprisingly, the smaller and/or
more concentrated markets found that an exchange did not work very well in their contexts.
Consequently, various liquidity-supporting measures were implemented to incentivize incum-
bent utilities and TSOs to trade on power exchanges for the benefit of new entrants, which
relied on them to survive. Note that TSOs only purchase energy to offset the losses in their
transmission network. Today, day-ahead markets in Europe are generally considered to be
liquid enough, while this is not yet the case for intraday markets.
Third, we discuss the circumstances that led to a consolidation of the number of power
exchanges across Europe. In the early days of the EU electricity market integration process,
every country wanted to have its own market infrastructure. However, after the initial sensitiv-
ities faded, the market logic led to consolidation. Often, these national power exchanges had
in any case already outsourced a major part of their activities instead of developing their own
trading software and/or platforms. Gradually, two large players emerged through mergers and
 Who gets the rights to trade across borders? 39

Figure 2A.1 European power exchange mergers and acquisitions (non-exhaustive)

acquisitions. One large player is Nord Pool AS, which is mentioned in Section 2.3. The other is
EPEX SPOT SE, which has a long history of mergers and collaborations, as is shown in Figure
2A.1. In 2002, the Dutch power exchange APX tried to set up a Belgian subsidiary called BPX,
a project which was abandoned. Instead, the Belgian TSO Elia established a Belgian power
exchange called Belpex in 2006. Belpex was eventually integrated into APX in 2010. Five
years later, APX itself was integrated into EPEX SPOT, which was the result of the merger of
the German exchange EEX and the French Powernext. At the time of writing, EPEX SPOT
and Nord Pool are active in 8 and 14 countries respectively, with more expansions planned for
both power exchanges. The south-eastern European countries are still setting up new power
exchanges in cooperation with one of the two big players. Examples are the Independent
Bulgarian Energy Exchange (IBEX) and the Croatian power exchange (CROPEX), where
Nord Pool operates the day-ahead and intraday markets. Another example is the South East
European Power Exchange (SEEPEX), which was established as a joint venture between the
Serbian TSO and EPEX SPOT. We also see new players coming into the market infrastructure
business, and will come back to this in Chapter 8.
40 Evolution of electricity markets in Europe

2A.2 ANNEX: A SIMPLE NUMERICAL EXAMPLE OF MARKET

COUPLING

Imagine two countries, A and B, which each represent one bidding zone. The concept of
bidding zones is discussed in more depth in Chapter 3 of this book. Basically, when a country
equals one bidding zone it means that the electricity price will be the same for the whole
country per market time period.
First, no later than 11.00 Central European Time (CET), available capacities on intercon-
nectors are published. Second, until 12.00 CET market parties have the possibility to submit
their buy/sell orders to the power exchange(s) in their country for the day-ahead auction
covering delivery in all hours of the next day. Third, each power exchange sends its order
books, that is, the collected buy and sell orders, to the market coupling operator (MCO). The
following table shows an example for an order book of two countries for a specific hour.
Fourth, the prices are calculated by the pan-European algorithm operated by the MCO. Fifth,
under normal circumstances, at 12.55 CET the final market coupling results are published by
the power exchange(s).
Country A Sell orders Inelastic demand
Ga1: 20 MWh at €20/MWh 80 MWh
Ga2: 30 MWh at €50/MWh
Ga3: 70 MWh at €60/MWh
Country B Sell orders Inelastic demand
Gb1: 130 MWh at €20/MWh 100 MWh
Gb2: 20 MWh at €30/MWh
Gb3: 40 MWh at €40/MWh

We can consider four cases, which differ in the amount of commercial cross-border capacity
between the countries indicated by the relevant TSOs. For simplicity, an inelastic demand is
considered, and the price is set by the marginal sell order unconditional on whether it is fully
filled or not.

Case 1: no commercial cross-border capacity between countries A and B

This means there are two different clearings, one for each country.
Price Demand Supply Export to other country
Country A €60/MWh 80 MWh 80 MWh 0
Country B €20/MWh 100 MWh 100 MWh 0

In this case, there is no cross-border trade. The spread between the two countries is €40/MWh.
In country A, Ga3 sets the price. In country B, Gb1 sets the price. There is no congestion rent
as there is no volume traded between the two countries.

Case 2: unlimited commercial cross-border capacity between countries A and B

This means there is one joint clearing for both countries, that is, the supply (and demand)
curves of the two countries are aggregated.
 Who gets the rights to trade across borders? 41

Price Demand Supply Export to other country

Country A €40/MWh 80 MWh 20 MWh − 60 MWh
Country B €40/MWh 100 MWh 160 MWh + 60 MWh

In this case, the price reduces in country A and increases in country B. There is cross-border
trade but no price spread. Gb3 sets the price for both countries. Implicitly, 60 MW of
cross-border capacity is allocated by the MCO from country B to A for the particular hour.
There is no congestion rent as there is no price spread between the two countries.

Case 3: 100 MW/h commercial cross-border capacity is available in both directions between
countries A and B

This means there is one joint clearing for both countries, that is, the supply (and demand)
curves of the two countries are aggregated unless the exchange exceeds the commercial
capacity available.
Price Demand Supply Export to other country
Country A €40/MWh 80 MWh 20 MWh − 60 MWh
Country B €40/MWh 100 MWh 160 MWh + 60 MWh

This case is no different to case 2 as the exchange between the two countries is not limited by
the commercial cross-border capacity available.

Case 4: 40 MW/h commercial cross-border capacity between countries A and B is available

in both directions

This means there is one joint clearing for both countries, that is, the supply (and demand)
curves of the two countries are aggregated unless the exchange exceeds the commercial
capacity available.
Price Demand Supply Export to other country
Country A €50/MWh 80 MWh 40 MWh − 40 MWh
Country B €30/MWh 100 MWh 140 MWh + 40 MWh

In this case, compared to case 1, the price reduces in country A (but not as much as in cases
2 and 3) and the price increases in country B (but not as much as in cases 2 and 3). There is
cross-border trade and a price spread. The commercial exchange between the countries is
limited due to the capacity available. In country A, Ga2 sets the price; in country B, Gb2 sets
the price. Implicitly, 40 MW of cross-border capacity is allocated from country B to A for the
particular hour by the MCO. There is a congestion rent of 40 MWh*€20/MWh = €800 for this
particular hour. More information about the calculation of congestion rent can be found in
Annex 4A.1 of Chapter 4.
42 Evolution of electricity markets in Europe

2A.3 ANNEX: REGULATORY GUIDE

Table 2A.1 Regulatory guide

Section of this chapter, topic and relevant regulation Relevant articles

Section 2.1
The First Directive 96/92/EC stated that TSOs had to provide Art. 7(5) states that ‘The system operator shall not discriminate
different network users with non-discriminatory access to their between system users or classes of system users, particularly in
networks. favour of its subsidiaries or shareholders.’
The First Directive 96/92/EC allowed Member States to ask In accordance with Art. 24, the Member States were entitled to
for a transitional exemption from the relevant article in the ask the Commission to allow an exemption up to one year after
legislation. the entry into force of the Directive.
Section 2.2
Regulation (EC) No 1228/2003 required a market-based The first point under ‘general’ in the annex to Regulation
approach for the allocation of transmission rights. (EC) No 1228/2003 states that ‘Congestion management
method(s) implemented by Member States shall deal with
short-run congestion in a market-based, economically efficient
manner whilst simultaneously providing signals or incentives
for efficient network and generation investment in the right
locations.’
Section 2.3
The CACM GL makes day-ahead market coupling binding for Art. 42(1) states that ‘The day-ahead cross-zonal capacity
all. charge shall reflect market congestion and shall amount to
the difference between the corresponding day-ahead clearing
prices of the relevant bidding zones.’
The operation of the algorithm is called the market coupling Art. 2(30) defines the MCO function as ‘the task of matching
operator (MCO) function in the CACM GL. orders from the day-ahead and intraday markets for different
bidding zones and simultaneously allocating cross-zonal
capacities.’
According to the CACM GL, to be able to participate in Art. 2(23) defines a NEMO as ‘an entity designated by
market coupling, exchanges have to be certified as Nominated the competent authority to perform tasks related to single
Electricity Market Operators (NEMOs). day-ahead or single intraday coupling.’
According to the CACM GL, the NEMO function is jointly Art. 7(2) states that ‘NEMOs shall carry out MCO functions
operated by all participating power exchanges. jointly with other NEMOs’and lists the different tasks.
The default option described in the CACM GL is to have The competitive model is the default (preferred) arrangement.
competing power exchanges in each Member State. See Art. 4(1, 5). However, a monopoly is still possible
according to Art. 4(6.a) and Art. 5(2). Art. 5(2) states
‘a national legal monopoly is deemed to exist where national
law expressly provides that no more than one entity within
a Member State or Member State bidding zone can carry out
day-ahead and intraday trading services.’ Art. 5(3) continues
‘if the Commission deems that there is no justification for the
 Who gets the rights to trade across borders? 43

Section of this chapter, topic and relevant regulation Relevant articles

continuation of national legal monopolies or for the continued
refusal of a Member State to allow cross-border trading by
a NEMO designated in another MS, the Commission may
consider appropriate legislative or other appropriate measures
to further increase competition and trade between and within
Member States.’
According to the CACM GL, the regulator (or other competent Art. 4(3): ‘Unless otherwise provided by Member States,
authority) designates the one or more power exchanges that regulatory authorities shall be the designating authority,
can organize cross-border trade in the day-ahead and intraday responsible for NEMO designation, monitoring of compliance
markets. with the designation criteria and, in the case of national legal
monopolies, the approval of NEMO fees or the methodology
to calculate NEMO fees. Member States may provide that
authorities other than the regulatory authorities be the
designating authority. In these circumstances Member States
shall ensure that the designating authority has the same rights
and obligations as the regulatory authorities in order to
effectively carry out its tasks.’
According to the CACM GL, competitive NEMOs designated This is described in Art. 4(5). On top of not having the right
in one Member State also have the right to offer trading to offer trading services in another Member State when the
services with delivery in another Member State where NEMO is a legal monopoly in its Member state or where
a competitive model is implemented, unless an exception is a legal monopoly is in place in the Member State of delivery,
justified. two other exceptions are listed in Art. 4(6): technical obstacles
and incompatible trading rules.
Multi-NEMO arrangements (MNAs) in the CACM GL. Arrangements concerning more than one NEMO in one bidding
zone are described in Art. 45 for day-ahead and Art. 57 for
intraday.
Operational MCO implementation according to the CACM GL. Art. 7(3) requires all NEMOs to come up with a plan that sets
out how to jointly set up and perform the MCO functions. The
MCO plan (All NEMOs 2017c) was approved by all NRAs on
26 June 2017.
Having a single independent MCO entity is a possible option Recital 15 states that ‘The Commission, in cooperation with
according to the CACM GL. the ACER may create or appoint a single regulated entity
to perform common MCO functions relating to the market
operation of single day-ahead and intraday coupling.’
The CACM GL states that TSOs may contribute to the Art. 76(1) states that the costs of establishing, amending and
MCO-function-related costs of the NEMOs concerned but are operating the algorithms, systems and procedures for market
not obliged to. coupling (day-ahead and intraday) are borne by all NEMOs.
Art. 76(2) adds that TSOs may contribute to the costs subject to
approval by the relevant regulatory authorities.
The CACM GL requires NEMOs to submit a joint proposal for Art. 40 and Art. 53 describe the methodology for the products
the bidding formats they will continue to use. to be accommodated in the SDAC and SIDC respectively.
For both market timeframes, the orders resulting from these
products should be expressed in euros and make reference to
one or multiple market time units. In January 2018, all NRAs
approved the amended proposals for SDAC and SIDC products
(All NEMOs 2017a, 2017b).
44 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant regulation Relevant articles

Side payments would require a change to the CACM GL. Art. 38(1.b) states that‘The price coupling algorithm shall
produce the results set out in Article 39(2), in a manner which:
… (b) uses the marginal pricing principle according to which
all accepted bids will have the same price per bidding zone per
market time unit.’
Regulation (EU) 2019/943 of the Clean Energy Package More specifically, in Art. 8(2) of Regulation (EU) 2019/943,
includes provisions that will increase the granularity of it is stated that ‘NEMOs shall provide market participants
day-ahead markets from hourly markets to half-hourly or even with the opportunity to trade in energy in time intervals which
15 minutes. are at least as short as the imbalance settlement period for
both day-ahead and intraday markets.’In the same Regulation
and article it is stated at point 4 that ‘By 1 January 2021,
the imbalance settlement period shall be 15 minutes in all
scheduling areas, unless regulatory authorities have granted
a derogation or an exemption. Derogations may be granted
only until 31 December 2024. From 1 January 2025, the
imbalance settlement period shall not exceed 30 minutes where
an exemption has been granted by all the regulatory authorities
within a synchronous area.’
Following the CACM GL, all NEMOs were asked to come up Arts. 41 and 54 describe the methodology for the harmonized
with a joint proposal for minimum and maximum prices in the minimum and maximum prices for the SDAC and SIDC
SDAC and SIDC. respectively. With decisions 04/2017 and 05/2017, ACER
(2017a, 2017b) adopted the final version of the harmonized
min-max clearing prices to be applied, respectively, to SDAC
and SIDC in November 2017.
Section 2.4
According to the FCA GL, JAO became the European platform Recital 5 states that ‘Harmonised long-term cross-zonal
for long-term transmission right allocation. capacity allocation rules require the establishment and
operation of a single allocation platform at European level.
This central platform should be developed by all TSOs to
facilitate the allocation of long-term transmission rights
for market participants and should provide for the transfer
of long-term transmission rights from one eligible market
participant to another.’Arts. 48, 49 and 50 respectively
describe the establishment, functional requirements and tasks
of the single allocation platform.
To split the available transmission rights over the different Art. 16 describes the methodology for splitting long-term
long-term timeframes and contract lengths, the FCA GL cross-zonal capacity.
foresees the need to develop methodologies at the regional
level.
 Who gets the rights to trade across borders? 45

Section of this chapter, topic and relevant regulation Relevant articles

In the FCA GL, two causes for curtailment of long-term First, the curtailment of transmission rights in the event of
transmission rights are distinguished. force majeure. A force majeure is defined in the CACM GL
Art. 45(2) as ‘any unforeseeable or unusual event or situation
beyond the reasonable control of a TSO, and not due to
a fault of the TSO, which cannot be avoided or overcome
with reasonable foresight and diligence, which cannot be
solved by measures which are from a technical, financial
or economic point of view reasonably possible for the TSO,
which has actually happened and is objectively verifiable, and
which makes it impossible for the TSO to fulfil, temporarily
or permanently, its obligations in accordance with this
Regulation.’ Second, long-term transmission rights can also be
curtailed prior to the day-ahead firmness deadline to ensure that
operation remains within operational security limits, defined in
CACM GL Art. 2(7) ‘as the acceptable operating boundaries
for secure grid operation such as thermal limits, voltage limits,
short-circuit current limits, frequency and dynamic stability
limits.’ Lastly, there can be a special case when bidding zone
borders cease to exist. This could happen when two bidding
zones are merged or bidding zone borders are redrawn.
According to the FCA GL, if what happens is considered Art. 56(3) states that in the case of force majeure, the holder
force majeure, the price of the right in the original auction is of long-term transmission rights will receive compensation
refunded. from the TSO which invoked the force majeure. This
compensation will be equal to the amount initially paid for
long-term transmission rights. Art. 27(2) states that similarly
in the special case that a bidding zone border ceases to
exists, the transmission right holders shall also be entitled to
reimbursement by the TSOs concerned based on the initial
price paid for the long-term transmission rights.
If it is not a force majeure, i.e. curtailment of the right to In the case of a curtailment due to operational security limits,
ensure operation remains within operational security limits, the Art. 53(2) specifies that the TSOs concerned on the bidding
compensation is the lost opportunity, which is the day-ahead zone border where long-term transmission rights have been
price spread. In this case, the TSOs concerned might propose curtailed shall compensate the holders of these rights with the
introducing a cap, which is further specified in the FCA GL. market spread. Furthermore, Art. 54(1) adds that the TSOs
concerned on a bidding zone border may propose a cap on
the total compensation to be paid to all holders of curtailed
long-term transmission rights.
The determination of whether an event is classified as force Art. 56(5) states that‘Where a Member State has so provided,
majeure is, however, still done at the national level. upon request by the TSO concerned, the national regulatory
authority shall assess whether an event qualifies as force
majeure.’
46 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant regulation Relevant articles

The FCA GL leaves it open whether PTRs or FTRs are used. Art. 31(1) states: ‘Long-term cross-zonal capacity shall be
If a trader decides not to use a PTR, the trader is compensated allocated to market participants by the allocation platform
for the value of the right in the day-ahead auction, where other in the form of physical transmission rights pursuant to the
traders might be willing to pay for it (use-it-or-sell-it). UIOSI principle or in the form of FTRs – options or FTRs
– obligations.’ Art. 2(6) defines UIOSI as ‘the principle
according to which the underlying cross-zonal capacity of
physical transmission rights purchased and non-nominated
is automatically made available for day-ahead capacity
allocation and according to which the holder of these physical
transmission rights receives remuneration from the TSOs.’
The FCA GL requires that marginal pricing is applied in the Art. 28(1) states that ‘The allocation of forward capacity
auctions. shall take place in a way which: (a) uses the marginal pricing
principle to generate results for each bidding zone border,
direction of utilization and market time unit.’
The FCA GL requires that for both PTRs and FTRs harmonized Art. 51 describes the introduction of harmonized allocation
allocation rules are followed. rules. Art. 52 describes the requirements for the harmonized
allocation rules. The revised proposal by all TSOs regarding
harmonized allocation rules was not unanimously approved by
all NRAs. Finally, ACER adopted a decision in August 2017.
The FCA GL states that the two types of transmission rights Art. 31(6) states that ‘The allocation of physical transmission
cannot be applied in parallel on one border. rights and FTRs – options in parallel at the same bidding zone
border is not allowed. The allocation of physical transmission
rights and FTRs – obligations in parallel at the same bidding
zone border is not allowed.’
Regulators can adopt a coordinated decision not to issue PTRs Art. 30(1) states that ‘TSOs on a bidding zone border shall
or FTRs on a border when they can show that there is no need issue long-term transmission rights unless the competent
for hedging or by ensuring that there are other cross-border regulatory authorities of the bidding zone border have adopted
hedging instruments available in the market. coordinated decisions not to issue long-term transmission
rights on the bidding zone border. When adopting their
decisions, the competent regulatory authorities of the bidding
zone border shall consult the regulatory authorities of the
relevant capacity calculation region and take due account
of their opinions.’ The remainder of Art. 30 states that an
assessment needs to show that no hedging needs are unmet by
ensuring that there are other cross-border hedging instruments
available in the market.
Section 2.5
The CACM GL prescribes that the intraday cross-border gate Art. 59(3) states that ‘one intraday cross-zonal gate closure
closure shall be at most one hour before delivery. After the time shall be established for each market time unit for a given
intraday cross-border gate closure, national intraday markets bidding zone border. It shall be at most one hour before
often remain open until even closer to real time. the start of the relevant market time unit and shall take
into account the relevant balancing processes in relation
to operational security.’The intraday cross-zonal gate
closure time is defined in Art. 2(3): ‘the point in time where
cross-zonal capacity allocation is no longer permitted for
a given market time unit.’
 Who gets the rights to trade across borders? 47

Section of this chapter, topic and relevant regulation Relevant articles

Continuous trade is the default trading arrangement in Art. 51 states that ‘From the intraday cross-zonal gate opening
the intraday timeframe, and auctions are tolerated as time until the intraday cross-zonal gate closure time, the
a complementary regional arrangement. continuous trading matching algorithm shall determine which
orders to select for matching.’ Art. 63(1) adds that the relevant
NEMOs and TSOs on bidding zone borders may jointly submit
a common proposal for the design and implementation of
complementary regional intraday auctions. Complementary
regional intraday auctions may be implemented within or
between bidding zones in addition to the single intraday
coupling solution referred to in Art. 51.
The CACM GL requires a single methodology for intraday Art. 55(3) states that all TSOs shall develop a proposal for
cross-zonal pricing reflecting market congestion through an a single methodology for pricing intraday cross-zonal capacity.
implicit allocation method. Art. 55(1) adds that ‘Once applied, the single methodology for
pricing intraday cross-zonal capacity developed in accordance
with Article 55(3) shall reflect market congestion and shall
be based on actual orders.’ The revised proposal by all TSOs
regarding a single methodology for pricing intraday cross-zonal
capacity was not unanimously approved by all NRAs. Finally,
ACER (2019a) adopted a decision in January 2019.
According to the CACM GL, explicit allocation is only allowed Art. 64 states the provisions related to explicit allocation under
as a transitional complementary arrangement. the section title ‘Transitional intraday arrangements’. In Art.
64(1) it is stated that ‘Where jointly requested by the regulatory
authorities of the Member States of each of the bidding zone
borders concerned, the TSOs concerned shall also provide
explicit allocation, in addition to implicit allocation, that is to
say, capacity allocation separate from the electricity trade, via
the capacity management module on bidding zone borders.’
Art. 65 explains the removal of explicit allocation.
3. How to calculate border trade constraints?
Leonardo Meeus with Tim Schittekatte

In this chapter we answer five questions. First, why do we focus so much on constraints?
Second, why do we calculate trade constraints on virtual borders? Third, who is best placed
to do virtual border calculations? Fourth, how to organize the data exchange in support of the
calculations? And fifth, why are there still open issues?

3.1 WHY DO WE FOCUS SO MUCH ON CONSTRAINTS?

In this section, we explain that transmission lines are currently the only economically viable
way to transport electricity, transmission planning is challenging and transmission operation
is critical to avoid blackouts.
First, transmission lines are currently the only economically viable way to transport elec-
tricity. In other industries, transport and logistics follow the rest of the value chain. There
are many companies that compete to supply transport services. In those industries, transport
is considered a cost rather than a constraint. Maybe in the future we will have competition
to transport electricity, with market parties carrying large-scale batteries on trucks, trains or
ships to transport it over long distances. However, for the moment there are no alternatives.
In the EU, we have transmission system operators (TSOs), which own networks and perform
the system operator tasks in a certain geographical area that is called a control area. In most
countries, the control area coincides with the national territory. Germany is one of the excep-
tions, having four control areas. Some countries have chosen to separate the system operator
role from the asset owner, in which case the system operator is referred to as an independent
system operator (ISO). For example, Ireland and Great Britain follow this model, which is
also the dominant one in the US. Regardless of the specific model, the transmission network
is centrally planned and regulated, and therefore by definition imperfect. In economic terms,
we consider the transmission network to be a natural monopoly. We discussed how incentive
regulation has been used to manage these imperfections in our previous book.1
Second, transmission planning is challenging. Transmission planning by TSOs (or ISOs)
under the supervision of regulators tries to anticipate the decisions of market parties on where
they will invest in power plants or where they will consume electricity. Investments in the
network are then planned accordingly. However, as there are many uncertainties involved
and large transmission projects take longer than most power plant projects, there are typically
many transmission-capacity constraints in the system. Due to these constraints, the cheapest
resources cannot reach the consumers with the highest willingness to pay. This planning
problem is becoming even more pressing with greater numbers of renewable energy technolo-
gies. Indeed, wind turbines and solar parks are faster to build than coal or nuclear power plants.

48
 How to calculate border trade constraints? 49

Third, transmission operation is critical to avoid blackouts. If we do not account for lim-
itations in the network, the market will result in consumption and production patterns that
are not feasible. If we then force the flows through the network lines, they overheat and dis-
connect. The weakest link disconnects first, which redistributes the flows to other lines. This
can create additional overheating, which is followed by a cascade of disconnections, ending
in a blackout. The same cascade of events can take place if something unexpected happens
to a transmission line. An example of such an event is the blackout that took place in Italy in
2003 after a line tripped on the Swiss–Italian border (Box 3.1). Normally, these problems are
avoided by applying the N−1 redundancy principle. There is always some reserve kept in the
system to avoid a complete collapse if one of the elements fails.

BOX 3.1 THE ITALIAN BLACKOUT OF 28 SEPTEMBER 2003

On 28 September 2003 at 03.28 in the morning, a blackout affected more than 56 million
people across Italy and some areas of Switzerland. The consequences were huge. For exam-
ple, 30 000 people were stranded on trains in the Italian railway network. Estimates of the
number of fatalities directly related to the loss of power vary. Much of Italy was energized
again before 08.00, the central part around noon and the rest of the mainland at 17.00. Sicily
was only fully energized at 21.40. Some commercial and domestic users even suffered dis-
ruption in their power supplies for up to 48 hours.
The sequence of events was triggered by the 380 kV Swiss Mettlen–Lavorgo line trip-
ping at 03.01 as a result of a tree flashover. Tree flashovers happen when a line comes too
close to trees because they have not been trimmed properly and/or when a line sags more
than normal due to overheating.
Following the N−1 principle, other lines were able to take over the flows, but only for
a short time. The system needed to be restored to return to a secure state. The Swiss TSO
therefore asked the Italian TSO to reduce Italian imports by 300 MW. At this time, Italy was
importing up to 300 MW more than the already high import schedule, which amounted to
6400 MW on the northern border. However, the import reduction, together with some in-
ternal countermeasures taken within the Swiss system, was insufficient to relieve the over-
load. At 03.25, the Sils–Soazza line in proximity to the Mettlen–Lavorgo line also tripped
after another tree flashover. Almost simultaneously and automatically the remaining lines
towards Italy also tripped, and the Italian system was isolated from the European network
about 12 seconds after the loss of the Sils–Soazza line.
After this separation, the Italian system had a large shortage resulting in a frequency drop
and a disconnection of power plants in the north of Italy. This then accelerated the frequen-
cy drop. The Italian TSO tried to save the situation with emergency load shedding, but the
frequency reached the threshold of 47.5 Hz and the whole system collapsed in a blackout
2 minutes and 30 seconds after the isolation of Italy from the rest of Europe. In Chapter 6
we will come back to the role of power plants and other grid connections in this kind of
incident.
50 Evolution of electricity markets in Europe

Note: UCTE (2004) provides a detailed investigation of the incident. Ranci (2019), the first president of the
Italian regulator and founding director of the Florence School of Regulation, describes from his point of view how
he saw the blackout that happened in the last months of his mandate. He states that it was an unexpected failure,
showing how limited the regulator’s powers were to coordinate international relations. He also comments that the
Italian regulatory authority had to learn how difficult it can be to explain a highly technical issue to the media, to
the public at large and even to members of government.

3.2 WHY DO WE CALCULATE TRADE CONSTRAINTS ON

VIRTUAL BORDERS?
In this section, we start by introducing the zonal congestion pricing approach that is currently
in place in the EU. Afterwards, we discuss why the calculation of net transfer capacities
between bidding zones is problematic. Last, we explain that flow-based market coupling
allows moving away from net transfer capacities.
First, the zonal congestion pricing approach. When electricity trading started in Europe, the
general view was that transmission constraints would not be an issue within a country. At the
national level, investments had been made to avoid constraints, while internationally a more
minimalistic approach had been followed. Therefore, it seemed logical to define trade borders
between countries – so-called bidding zones – and allocate a limited amount of transmission
rights on these borders, so-called net transfer capacities. This approach is referred to as
a bidding-zone market configuration. This implies that prices are zonal, meaning that within
a predefined bidding zone the wholesale electricity prices at each market time unit are the
same and trade is unlimited. Of course, everybody knew that bidding zone borders were virtual
because a border is an aggregation of physical lines. In some cases, there was awareness that
the actual constraint was not necessarily the physical lines crossing the border but another
network component deeper inside the national network that was strongly impacted by an
increase in cross-border trade. However, this initially did not lead to questioning the approach.
Figure 3.1 shows the bidding zone configuration in Europe in 2019. Bidding zone borders still
mostly coincide with national borders. The exceptions are Sweden, Norway, Denmark and
Italy, which are divided into several bidding zones, and Luxembourg, which shares a bidding
zone with Germany.
Second, the calculation of net transfer capacity (NTC). TSOs had the impossible task of
calculating the amount of energy that could be traded over a virtual border. First, the amount
depends on how the flow will be distributed over the different border lines, which in turn
depends on production and consumption patterns, which are only known after trade takes
place. TSOs therefore predict the outcome of trade when deciding how much capacity to
make available for trade. Second, it depends on what happens on other borders. The amount
of capacity that a certain TSO can make available on one border depends on what other TSOs
are doing on other borders. There are many flow paths between two points in a transmission
network, and the flow will automatically distribute itself over all the paths. As a result, TSOs
were very conservative in estimating NTC values to avoid getting it wrong and having to pay
penalties for not delivering the capacity they had made available to the market.
Third, the evolution towards flow-based market coupling (FBMC). FBMC essentially
means that now virtual border capacity and virtual flow distribution factors are calculated.
By doing this, TSOs can be less conservative in the amount of capacity they make available.
 How to calculate border trade constraints? 51

Figure 3.1 Bidding zones in Europe as of November 2019

This is because the interdependencies between different bidding zones are considered using
virtual flow factors in the market coupling algorithm. The algorithm allocates the interde-
pendent flows over the different borders to maximize welfare from cross-border trade. The
Capacity Allocation and Congestion Management Guideline (CACM GL) refers to FBMC as
the primary approach for day-ahead and intraday capacity calculations. Even though the NTC
method is still allowed, FBMC is already the dominant model. The idea was discussed long
ago in the central western Europe (CWE) regional initiative. The original launch in that region
was foreseen for 2009, but due to delays and implementation issues it took until the adoption
of the CACM GL in 2015 to finally implement the FBMC method. For a more detailed discus-
sion of the FBMC concept and the key parameters that play a role, see Annex 3A.1.2

3.3 WHO IS BEST PLACED TO DO VIRTUAL BORDER

CALCULATIONS?

Regional collaboration among TSOs started voluntarily and was then formalized. In what
follows, we first discuss the steps from regional security coordination initiatives (RSCIs) to
regional security coordinators (RSCs) to finally regional coordination centres (RCCs). After,
52 Evolution of electricity markets in Europe

we discuss the connection of RCCs with capacity calculation regions (CCRs), which play
a key role in the implementation of FBMC.
First, the steps from RSCIs to RSCs to RCCs. Around 2008, the first RSCIs set up were
CORESO (based in Brussels) and TSCNET services (based in Munich). In 2015, another
RSCI was created in south-eastern Europe (SEE) in Belgrade. In 2016, Nordic and Baltic
RSCIs were established. A system operation guideline (SO GL) adopted in 2017 formalized
these initiatives by stating that each control area was to be covered by at least one RSC. The
left-hand side of Figure 3.2 gives an overview of the geographical coverage of the five RSCs
as of January 2019. RSCs are owned or controlled by TSOs and perform different tasks. The
SO GL defines five core tasks for the RSCs, and the Clean Energy Package (CEP), more
specifically Regulation (EU) 2019/943, has recently added additional tasks. The CEP also
renamed the RSCs as regional coordination centres and enhanced their governance.
Second, the connection between RCCs and the CCRs. Within a CCR there is a high degree
of interdependence of capacity calculation across borders. CCRs play a key role in the
implementation of FBMC. The CACM GL includes the methodology for defining CCRs and
requires a coordinated capacity calculation methodology to be developed per CCR. It is this
capacity calculation methodology that specifies the implementation of FBMC using parame-
ters, which we discuss in Annex 3A.1. The definition of the geographical scope of the CCRs
was a very sensitive topic, with the Agency for the Cooperation of Energy Regulators (ACER)
deciding to merge central western Europe (CWE) and central eastern Europe (CEE) into one
region, referred to as the Core region. Remember that ACER can only decide on the adoption
of methodologies if the national regulatory authorities (NRAs) cannot agree, which was the
case for this methodology. For each CCR, the relevant RCC is appointed as coordinated
capacity calculator. In the case that there is more than one established RCC active in a CCR,
it is expected that one RCC will be responsible for assuming the function of the coordinated
capacity calculator on a rotating basis. The right-hand side of Figure 3.2 shows the resulting
map of the CCRs. Two of the ten CCRs, the Core and the Nordic CCR, agreed to implement
FBMC for the day-ahead timeframe. FBMC spanning the full Core region is expected to go
live on 1 December 2020. For the Nordics, the go-live is planned for 1 July 2021. The idea is
that gradually CCRs will merge and as such coordinated capacity calculation can be performed
over wider geographical areas.
Note finally that even though RCCs take up the role of coordinated capacity calculators,
TSOs remain responsible for maintaining operational security. TSOs can decide not to imple-
ment a coordinated action proposed by the RCC, whether it is an action related to capacity cal-
culation or grid operation, but only for security reasons, and not for economic considerations.
They then need to report the detailed reason to the RCC and the TSOs of the region. This will
be further investigated. The RCC may then propose a different set of actions. How this will
work is an open issue. In Chapter 6, we describe the incident that led to the start of RSCIs and
the tasks they and their grandchildren perform.

3.4 HOW TO ORGANIZE THE DATA EXCHANGE IN SUPPORT

OF THE CALCULATIONS?

To understand how the data exchange supporting regional capacity calculations is organized,
it is necessary to read multiple EU network codes and guidelines. The CACM GL, Forward
 How to calculate border trade constraints? 53

Source: Both figures are adapted from ENTSO-E (2019).

Figure 3.2 Left: the five regional security coordinators (RSCs) as of 1 January 2019;
right: the ten capacity calculation regions (CCRs)

Capacity Allocation Guideline (FCA GL) and System Operation Guideline (SO GL) each
contain methodologies that when combined explain how the data exchange is organized. All
these methodologies have already been approved by the NRAs and adopted. In what follows,
we provide an overview of who provides the input data to the TSOs, the process for integrating
the data from the TSOs into a common grid model (CGM), how this CGM is used and the
physical infrastructure used to exchange these data.
The first question is who provides the input data to the TSOs? The CACM GL and FCA GL
each include generation and load data provision methodologies (GLDPM). These set out the
generation and load data that TSOs can request from grid users. This applies mainly to the grid
users connected at voltages of 220 kV or above. It can also apply to voltages below 220 kV if
they are used in regional operational security analysis of the capacity calculation timeframe
concerned. These data provisions apply regardless of whether the units are operated by the
TSO or a distribution system operator (DSO), including closed distribution systems (a type of
private network). TSOs can only use the GLDPM to request data if those data are not already
available to them through other national or EU legislation (e.g. the ENTSO-E Transparency
Platform), which can help to avoid double reporting. In addition, the TSOs can only use infor-
mation received under the GLDPM for capacity calculations.
The second question is what is the process for integrating the data from the TSOs into
a common grid model? Figure 3.3 summarizes the process. It starts with the TSOs’ individual
grid models (IGMs). The IGMs cover the power system characteristics, such as generation,
load and grid topology. They are scenarios with a forecast representation of the power system
for a given timeframe. The RCCs then merge the IGMs into CGMs, one for each timeframe.
The merging process is a continuous cycle updating the information available on the grid situ-
54 Evolution of electricity markets in Europe

Source: Adapted from ENTSO-E (2018a).

Figure 3.3 The common grid model (CGM) merging process

ation to ensure that each TSO has enough information for its area and each RCC has the neces-
sary pan-European view to do the capacity calculations for the region. Note that the CGMs are
pan-European, even though the capacity calculation is done regionally. It is expected that the
RCCs will take turns to merge the IGMs into CGMs. Note too that ENTSO-E stores the IGMs
and the CGMs in a portal called the pan-European Operational Planning Data Environment
(OPDE).
The third question is how is the CGM used? The SO GL, the CACM GL and the FCA GL
each require the establishment of CGMs. The difference between these different models lies
in the timeframes for which they are calculated and used. Following the CACM GL, the RCCs
must use the CGMs for their capacity calculations in the day-ahead and intraday timeframes.
Following the FCA GL, the RCCs can use the CGMs for their capacity calculations in the
timeframes before day-ahead, but they can also opt for a statistical approach. The SO GL
refers to the use of CGMs for various system operation processes.
The fourth question is what is the physical infrastructure used to exchange these data?
A new physical infrastructure called the Physical Communication Network (PCN) will be built
for the purpose of exchanging CGM-related data. In 2017, the TSOs decided to merge the PCN
with their existing network to exchange real-time data (the Electronic Highway). Both types of
data exchange are equally critical for TSOs to perform their tasks.
In conclusion, the days when each TSO was using its own grid model to perform capacity
calculations are behind us. There is widespread agreement on the benefits of a common
approach to reducing costs and improving performance. The organization of data exchanges to
support the ongoing market integration process is a topic that is being increasingly discussed.
One of several open issues is the data exchange between the TSO and DSOs and access
rights to different grid users’ data. In this regard, the key organizational requirements, roles
and responsibilities (KORRR) methodology is relevant. The KORRR methodology, which
is part of the SO GL, covers data needed beyond capacity calculation – for example, data
for real-time operations – and it complements the GLDPMs. The approved version of this
 How to calculate border trade constraints? 55

methodology states that it is to be decided at the national level whether distribution-connected

significant grid users (SGUs) have to provide their data directly to the TSO or through the
connecting DSOs.3

3.5 WHY ARE THERE STILL OPEN ISSUES?

FBMC was expected to improve the availability of transmission capacity thanks to a more
sophisticated capacity calculation. To monitor if the approach delivered, ACER and the
Council of European Energy Regulators (CEER) developed the concept of a benchmark
capacity, which is the capacity they expect TSOs to make available on a certain border. In
the first full year of FBMC in 2016, on average only 59 per cent of this capacity was made
available in central western Europe. This was slightly better than what was available before
the implementation of FBMC, but there was less improvement than expected. ACER also
estimated that we are losing several billions of euros a year by not reaching the benchmark
capacity.4
The reason is that FBMC does not address today’s most pressing issue, which is that the real
constraints in the network are no longer at the outdated virtual borders but within the bidding
zones. In what follows, we discuss how Directorate-General for Competition (DG COMP –
the European competition authority), the EU network codes and guidelines, and the EU Clean
Energy Package (CEP) have started to address the issue. Finally, we discuss the US approach
to this problem.
First, the approach by DG COMP. This started with an antitrust case against the Swedish
TSO Svenska Kraftnät (SvK) and continued with a case against the German TSO TenneT DE.
The SvK case started in 2006 after a complaint by the Danish energy industry association that
it suffered high prices because the Swedish TSO was blocking cheap imports from Norway
to Denmark via Sweden. DG COMP concluded that SvK had indeed constrained cross-border
trade to reduce internal congestion problems. This was found to be an abuse of the TSO’s
dominant position and discrimination against cross-border flows in favour of internal flows.
SvK initially stated that it would remedy the situation by investing in the constrained internal
transmission lines, but this was not enough for the Commission. To avoid being fined, in 2010
SvK agreed to split Sweden into four bidding zones. This was the best solution because traders
in Sweden now compete on a level playing field with cross-border traders for the limited inter-
nal capacity that is available in Sweden. When Sweden was one bidding zone, national trade
was unlimited and cross-border trade could only use the national trade’s leftovers.5
The TenneT DE investigation started in 2018. In this case, the complaint was that TenneT
DE limited the export of cheap renewable wind energy from Denmark to Germany. This
TSO also agreed to remedies to avoid being fined by DG COMP. TenneT DE guaranteed it
would make significantly more capacity available on the Danish border a few months after
the agreement with DG COMP, and would further increase capacity by 2026, when additional
investment projects will have been completed. This solution is not a good one, but it currently
seems to be the only politically feasible solution in Germany. Germany does not want to split
up the national bidding zone into smaller bidding zones to deal with internal constraints. Its
long-term solution is transmission investment, and in the meantime it has agreed to allow
more cross-border trade, although it knows that it does not have enough internal transmission
capacity to handle it. The German TSOs therefore intervene in the market with so-called redis-
56 Evolution of electricity markets in Europe

patching actions. Germany is basically changing the market outcome to prevent its internal
lines from overloading. These interventions are costly, and they also lead to market distortions.
The result is paying those that caused the problem to solve the problem. They can then create
more problems and earn more money. ACER estimated that in 2017 redispatch costs in the EU
had already reached 2 billion euros.6
Second is the approach followed by the CACM GL. This introduced a bidding zone review
process. The first bidding zone review was requested by ACER in 2016 for the central Europe
region, and ENTSO-E published the results in 2018. Four alternative bidding zone configura-
tions were assessed using a multi-criteria analysis. Following the CACM GL, the three main
criteria considered were the stability and robustness of the bidding zones, network security and
overall market efficiency. No alternative bidding zone configuration was found to outperform
the status quo on all the criteria. The review therefore proposed keeping the current configu-
ration. The TSOs did not feel comfortable carrying out this politically sensitive review, and
together with most observers we have been disappointed with the outcome. This explains why
this issue has again been picked up in the CEP.
Third is the approach followed by the CEP. Regulation (EU) 2019/943 introduced an
improved bidding zone review process. The implementation of this new process has already
started with the submission of a bidding zone review methodology and alternative configu-
rations by TSOs. All the NRAs then have three months to reach a unanimous decision on the
proposal. If they fail to reach a decision, then ACER will decide within an additional three
months. The TSOs will do the bidding zone review and analyse alternative configurations
using the agreed-upon methodology. They will then provide the relevant Member States
or their designated competent authorities with a joint proposal to amend or maintain the
current bidding zone configuration. This process is similar to that described in the CACM
GL. However, the end of the process is new. If Member States that opt to change the existing
bidding zone configuration cannot reach a unanimous decision, as a last resort the European
Commission will adopt a decision to amend or maintain it.
The European Commission had originally proposed legislation that would put it in charge
of configurating bidding zones. However, the European Council (representing the Member
States) opposed this and compromised with the European Parliament on what came to be
known as one of the most controversial elements in the CEP. In the final version of Regulation
(EU) 2019/943, the European Commission only has last resort powers when it comes to the
bidding zone configuration, but Member States have to comply with a transmission capacity
availability threshold. Regulation (EU) 2019/943 states that the available transmission capacity
of all transmission elements possibly limiting the amount of electricity that can be exchanged
between different bidding zones should be at least 70 per cent of a benchmark value. A process
has been set in motion to assess the extent to which different countries already comply with
this threshold and how they intend to comply by 2025. Basically, countries need to choose
between the two DG COMP cases – whether to follow the Swedish or the German approach to
stopping discrimination between cross-border and national trade.
Fourth is the US approach to this issue, that is, nodal pricing. If it is not possible to agree
on bidding zones, they could simply be abolished or reduced to the physical nodes in the
transmission network. This would mean that, instead of a simplified network with virtual
borders and flow factors, the physical constraints of the transmission network are represented
in the market coupling algorithm. If none of the constraints are binding and thermal losses are
 How to calculate border trade constraints? 57

excluded, there would be a single price for the whole of Europe. If one of them is binding,
the result would be a ‘weather map’ in which every point or node in the transmission network
would have a different price depending on where it is in relation to that network constraint.
The price map would change with weather conditions as the availability of wind and solar is
becoming an important factor in electricity production. In addition, temperature influences
electricity consumption. US markets have experience with nodal pricing in smaller-scale
markets; Europe could be the first to do it on a much larger international scale. We shall see
how this open issue plays out.7

3.6 CONCLUSION

In this chapter on the calculation of cross-border trade constraints, we have answered five
questions.
First, why do we focus so much on constraints? Transmission lines are currently the only
economically viable way to transport electricity. Transmission planning is challenging and
transmission operation is critical to avoid blackouts. The N−1 redundancy principle means
that there is always some reserve kept in the system to avoid a complete collapse if one of the
elements fails.
Second, why do we calculate trade constraints on virtual borders? It seemed natural to
define virtual trade borders between countries and assign a limited capacity to these borders.
The FBMC algorithm captures the interdependencies between borders and the CACM GL
mandates it to be the primary approach for day-ahead and intraday capacity calculation.
Third, who is best placed to do these virtual border calculations? Collaboration among
TSOs started voluntarily and evolved into being more formal. The CACM GL introduced the
concepts of CCRs and RSCs and the Clean Energy Package added responsibilities for RSCs
and renamed them as regional coordination centres (RCCs).
Fourth, how to organize the data exchange in support of the calculations? The SO GL, the
CACM GL and the FCA GL require the establishment of CGMs. They ensure that each TSO
has information to manage its area and each regional coordination centre has the pan-European
view. Interestingly, a dedicated European physical communication infrastructure supporting
multiple real-time and non-real-time services has been built up to support the data exchange.
And fifth, why are there still open issues? The current bidding zone configuration is out-
dated. Too many countries have a national bidding zone with internal congestion within it.
They therefore discriminate against cross-border trade in favour of national trade. The best
solution is to split bidding zones, but some countries prefer to keep their national bidding zone
despite the very high cost.

NOTES
1. In Meeus and Glachant (2018) we discuss how TSOs and distribution systems operators (DSOs)
are regulated as monopolies over transmission and distribution assets. We also show that the border
between what is considered a market, transmission or distribution is not always clear, which is
referred to as seams issues or grey areas in regulation.
2. A comprehensive comparison between the NTC approach and FBMC can be found in Van den
Bergh et al. (2016).
3. A distribution-connected significant grid user (SGU) is a generator connected to a distribution
network with a connected capacity higher than a certain threshold. This threshold, typically between
58 Evolution of electricity markets in Europe

several hundred kWs and 1 MW, depends on the national implementation of the Requirements for
Generators Network Code (RfG NC), which is discussed in more depth in Chapter 6. In addition,
distribution-connected demand facilities, directly or aggregated, which provide demand response to
the TSO are considered SGUs.
4. The benchmark capacity is calculated as if the ACER (2016b) recommendation on common
capacity calculation, redispatching and countertrading cost sharing methodologies is applied. This
recommendation emphasizes that internal network elements should not limit available tradable
cross-zonal capacity and that loop flows and other unscheduled flows should be minimized. Loop
flows are explained in Annex 3A.2. Making comparisons with how much available capacity was
offered in the past is not straightforward. This is first because it is not easy to compare available
tradable capacity under an NTC and an FBMC approach, and second because the concept of
a benchmark capacity was introduced for the first time in the ACER and CEER (2017) market
monitoring report of 2016. For example, the market monitoring analysis in ACER and CEER (2016)
shows that approximately 31 per cent of the interconnector’s physical capacity on CWE borders was
used for trading in 2015. The same report found that the ratio between NTC and thermal capacity in
CWE was 27 per cent.
5. A more detailed legal discussion of the SvK case can be found in the work of our colleagues de
Hauteclocque and Hancher (2011).
6. The discussion regarding the TenneT DE case is based on press releases from the European
Commission (2018). The €2 billion estimate of EU-wide redispatch costs in 2017 is based on the
ACER and CEER (2018) market monitoring report. Using simple cases, Stoft (1999) shows the
possibility of gaming between the wholesale market and redispatching, the so-called incremental–
decremental (inc–dec) game. Purchala (2019) describes how the zonal market does not facilitate
correct incentives for efficient behaviour.
7. The book Spot Pricing of Electricity by Schweppe et al. (1988) and the paper ‘Contract networks
for electric power transmission’ by Hogan (1992) were seminal works which laid the foundations
for the implementation of nodal pricing in the US. Neuhoff and many colleagues (2013) quantify
the Europe-wide operational savings of nodal relative to zonal market design to be in the order
of €0.8–2 billion/year (representing 1.1–3.6 per cent of operating costs). Similarly, Green (2007)
calculates that moving from a uniform zonal price to optimal nodal prices for England and Wales
could raise welfare by 1.3 per cent of the generators’ revenues, and would be less vulnerable to
market power. Both papers add that nodal pricing would send better investment signals but would
also cause politically sensitive redistributions. More recently, the concept of distribution locational
marginal pricing (DLMP) or nodal pricing all the way down to distribution grids is gaining atten-
tion. Caramanis et al. (2016) propose a DLMP framework in which the market and reserves are
cleared simultaneously (co-optimization) with inclusion of flexible loads and distributed energy
resources (DERs) and while considering line-flow constraints and distribution-bus voltage limits.
Papavasiliou (2018) provides an analysis of three DLMP approaches. He discusses price formation
under each approach and their strengths and limitations. MIT Energy Initiative (2016) highlights
the need for advanced meters, inexpensive information and communications technology (ICT)
solutions and DERs in order to make effective use of DLMP.

REFERENCES
ACER (2016a), ‘Decision No 06/2016 of the Agency for the Cooperation of Energy Regulators of 17
November 2016 on the Electricity Transmission System Operators’ Proposals for the Determination
of Capacity Calculation Regions’.
ACER (2016b), ‘Recommendation No 02/2016 of the Agency for the Cooperation of Energy Regulators
of 11 November 2016 on the Common Capacity Calculation and Redispatching and Countertrading
Cost Sharing Methodologies’.
ACER and CEER (2016), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2015 – Electricity Wholesale Markets Volume’.
ACER and CEER (2017), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2016 – Electricity Wholesale Markets Volume’.
 How to calculate border trade constraints? 59

ACER and CEER (2018), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2017 – Electricity Wholesale Markets Volume’.
ACER and CEER (2019), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2018 – Electricity Wholesale Markets Volume’.
All TSOs (2016), ‘All TSOs’ Proposal for a Generation and Load Data Provision Methodology in
Accordance with Article 16 of the CACM GL’, published on 13 May 2016.
All TSOs (2018), ‘All TSOs’ Proposal for the Key Organisational Requirements, Roles and
Responsibilities (KORRR) Relating to Data Exchange in Accordance with Article 40(6) of the SO
GL’, published on 1 October 2018.
Caramanis, M. C., E. Ntakou, W. W. Hogan, A. Chakrabortty and J. Schoene (2016), ‘Co-optimization of
power and reserves in dynamic T&D power markets with nondispatchable renewable generation and
distributed energy resources’, Proceedings of the IEEE, 104 (4), 807–36.
de Hauteclocque, A. and L. Hancher (2011), ‘The Svenska Kraftnät case: Introduction of bidding zones
in Sweden’, Network Industries Quarterly, 13 (1), 20–22.
ENTSO-E (2018a), ‘ENTSO-E implementation of the CIM’, presentation at CIM User Group on 7 June
2018, accessed at http://​cimug​.mobi/​Meetings/​eu2018/​2018 Ljubljana Presentations/CIM Meeting
Day 2/CIM ug ENTSO-E implementation of the CIM.pptx.
ENTSO-E (2018b), ‘First Edition of the Bidding Zone Review: Final Report’.
ENTSO-E (2019), ‘Enhanced TSO Regional Coordination for Europe’.
European Commission (2018), ‘Antitrust: Commission opens investigation into German grid operator
TenneT for limiting cross border electricity capacity with Denmark’, press release of 19 March 2018.
Green, R. (2007), ‘Nodal pricing of electricity: How much does it cost to get it wrong?’, Journal of
Regulatory Economics, 31 (2), 125–49.
Hogan, W. W. (1992), ‘Contract networks for electric power transmission’, Journal of Regulatory
Economics, 4, 211–42.
Meeus, L. and J.-M. Glachant (eds) (2018), Electricity Network Regulation in the EU: The Challenges
Ahead for Transmission and Distribution, Cheltenham, UK and Northampton, MA, USA: Edward
Elgar Publishing.
MIT Energy Initiative (2016), ‘Utility of the Future: An MIT Energy Initiative Response to an Industry
in Transition’.
Neuhoff, K., J. Barquin, J. W. Bialek, R. Boyd, C. J. Dent, F. Echavarren, T. Grau, C. von Hirschhausen,
B. F. Hobbs, F. Kunz, C. Nabe, G. Papaefthymiou, C. Weber and H. Weigt (2013), ‘Renewable
electric energy integration: Quantifying the value of design of markets for international transmission
capacity’, Energy Economics, 40, 760–72.
Papavasiliou, A. (2018), ‘Analysis of distribution locational marginal prices’, IEEE Transactions on
Smart Grid, 9 (5), 4872–82.
Purchala, K. (2019), ‘EU electricity market: The good, the bad and the ugly’, in S. Nies (ed.), The
European Energy Transition: Actors, Factors, Sectors, Deventer, Netherlands and Leuven, Belgium:
Claeys & Casteels, pp. 275–88.
Ranci, P. (2019), ‘The early years of the energy regulator in Italy’, Competition and Regulation in
Network Industries, 20 (3), 250–7.
Schweppe, F. C., M. C. Caramanis, R. D. Tabors and R. E. Bohn (1988), Spot Pricing of Electricity,
Boston, MA: Kluwer Academic Publishers.
Stoft, S. (1999), ‘Using game theory to study market power in simple networks’, IEEE Tutorial on Game
Theory in Electric Power Markets, 33–40.
UCTE (2004), ‘Final Report of the Investigation Committee on the 28 September 2003 Blackout in
Italy’.
Van den Bergh, K., J. Boury and E. Delarue (2016), ‘The flow-based market coupling in central western
Europe: Concepts and definitions’, Electricity Journal, 29 (1), 24–9.
60 Evolution of electricity markets in Europe

3A.1 ANNEX: FLOW-BASED MARKET COUPLING (FBMC)

FBMC is a process, not a one-step calculation, which starts two days before real time (base
case) and ends the morning one day ahead. At that moment, the coordinated capacity calcula-
tors deliver the necessary parameters to the market coupling operator (MCO) that is in charge
of the day-ahead market-clearing algorithm. In FBMC, market clearing and the allocation of
cross-zonal transmission capacity are done jointly. By considering interdependencies between
network elements in the market clearing, the transmission capacity is allocated in a way that
total welfare from trade is maximized. To arrive at a simplified network model without having
to consider all the individual lines, each TSO defines critical network elements (CNEs) in its
control area. In the literature, CNEs are also called critical branches. CNEs include cross-zonal
lines, but they can also include internal lines or transformers that are significantly impacted by
cross-zonal exchanges. By considering a CNE in the capacity calculation process, a CNE can
limit the amount of power that can be exchanged. The flow-based parameters incorporated in
the market-clearing algorithm are challenging to determine. We briefly introduce two of these:
zonal power transfer distribution factors (PTDFs) and CNE availability margins.
First, zonal PTDFs describe the linear relationship between the physical flow in a CNE and
the net exchange position of a specific bidding zone. The net exchange position is equal to the
total production minus the total consumption within the zone. If production is higher than con-
sumption, the zone will be a net exporter; vice versa, the zone would be a net importer. PTDFs
are formulated in a matrix with all the bidding zones in one dimension and all the CNEs in the
other dimension. With zonal pricing, nodes are grouped in zones. This implies that in order
to correctly represent flows between zones, zonal PTDFs need to be approximated from what
happens at the nodal level within a zone. Generation shift keys (GSKs) and load shift keys
(LSKs) may also be used to ‘translate’ changes in generation/consumption at the nodal level to
impacts on the net exchange level of a zone. GSKs are not easy to determine as they are based
on predictions of the market outcome and subject to forecast errors. Each TSO calculates the
GSKs for its control area.
Second, CNE available margins (AMs). The AM is the maximum flow from day-ahead
trade that can be carried by CNEs. The available margin is also referred to as the remaining
available margin (RAM). Determining the AM involves identifying the CNEs within a control
zone and the contingencies (critical outages) to be considered for system security. More pre-
cisely, AMs are the maximum flow an element can carry corrected by three types of flows to
what is left for day-ahead trade. The first correction is for the reference flow to a CNE. This
is caused by transactions between or within bidding zones other than the day-ahead market,
such as bilateral transactions or transactions in forward markets. The second correction is for
a final adjustment value, a margin which is TSO-specific and depends on, for example, costly
remedial actions such as redispatch possibilities. To determinate this value it is important to
ensure that there is no discrimination between internal and cross-zonal flows. The third cor-
rection is for a flow reliability margin, a safety margin compensating for approximations made
in the flow-based approach.
 How to calculate border trade constraints? 61

3A.2 ANNEX: TRANSIT FLOWS AND LOOP FLOWS

Two types of flow can ‘consume’ cross-zonal capacity between two bidding zones at the
expense of capacity available for trade between the two bidding zones. First, transit flows.
Some cross-zonal capacity of one bidding zone will be used by parallel flows resulting
from trade between other bidding zones. For example, trade between Germany and France
can flow through Belgium. Thus, the cross-zonal transmission capacity available for trade
between Belgium and its neighbours will be impacted. Transit flows are unscheduled when
the exchange causing the flow is cross-zonal and the capacity calculation is not coordinated
with the zone facing the flow. A flow is called an unscheduled flow when the physical flow
in the network differs from commercial exchanges between consumers and producers within
a bidding zone or between two different bidding zones. If the cross-zonal capacity calculation
is coordinated between the zones causing the transit flow and the zone facing it, the transit
flow is scheduled, as it can be accounted for in the transmission calculation process. A transit
flow is illustrated on the left-hand side of Figure 3A.1.
Second, loop flows. Transactions within a bidding zone can have an impact on the flows
through adjacent bidding zones. For example, if there is a commercial transaction between
the north and the south of Germany, it is possible for electricity to flow through Poland to
reach its destination. Thus, the cross-zonal transmission capacity available for trade between
Poland and its neighbours (mostly Germany in this case) will be impacted. By definition loop
flows are always unscheduled as they occur between two nodes within the same bidding zone.
A loop flow is illustrated on the right-hand side of Figure 3A.1.
Transit flows are exactly the kind of flows that can be dealt with by FBMC, but loop flows
cannot. Loop flows can only be addressed by revising the bidding zone configuration.

Figure 3A.1 Scheduled flows (black), transit flows (white, left) and loop flows (white,
right); the different rectangular areas represent bidding zones
62 Evolution of electricity markets in Europe

3A.3 ANNEX: REGULATORY GUIDE

Table 3A.1 Regulatory guide

Section of this chapter, topic and relevant Relevant article

regulation
Section 3.2
The CACM GL refers to FBMC as the primary Art. 20(1) states that ‘For the day-ahead market time-frame and intraday
approach for day-ahead and intraday capacity market time-frame the approach used in the common capacity calculation
calculations. Even though the NTC method is methodologies shall be a flow-based approach, except where the requirement
still allowed, FBMC is already the dominant under paragraph 7 is met.’Art. 20(7) states that ‘TSOs may jointly request the
model. competent regulatory authorities to apply the coordinated net transmission
capacity approach in regions and bidding zone borders (other than those
referred to in paragraphs 2 to 4), if the TSOs concerned are able to
demonstrate that the application of the capacity calculation methodology
using the flow-based approach would not yet be more efficient compared to
the coordinated net transmission capacity approach and assuming the same
level of operational security in the concerned region.’
Section 3.3
The SO GL formalizes RSCIs by stating that Art. 3(89) defines RSCs as ‘the entity or entities, owned or controlled by
each control area shall be covered by at least one TSOs, in one or more capacity calculation regions performing tasks related
regional security coordinator (RSC). to TSO regional coordination.’Art. 77 describes the organization of regional
operational security coordination. Art. 77(2.a) states that each TSO shall be
covered by at least one RSC.
Regulation (EU) 2019/943, part of the CEP, Art. 35 in Regulation (EU) 2019/943 describes the establishment and mission
renamed RSCs as regional coordination centres of RCCs. Art. 35(2) specifies ‘Following approval by regulatory authorities
(RCCs) and enhanced their governance. of the proposal [for the establishment of regional coordination centres], the
regional coordination centres shall replace the regional security coordinators
established pursuant to [SO GL] and shall enter into operation by 1 July
2022.’
The SO GL defines five core tasks of the RSCs. Art. 77(3) lists four tasks of the RSCs: (a) regional operational security
coordination; (b) building a common grid model; (c) regional outage
coordination; and (d) regional adequacy assessment. The fifth task,
coordinated capacity calculation, is indirectly mentioned in the CACM GL.
In recital 6 of the CACM GL it is stated that ‘capacity calculation for the
day-ahead and intraday should be coordinated at least at regional level.’
Additionally, Art. 6(3) of the Emergency and Restoration Network Code (ER
NC) states that RSCs will be consulted to assess the consistency of measures
described in a TSO’s system defence and restoration plan. Last, RSCs also
provide critical grid situation support to TSOs, a responsibility that was
introduced with the cold spell in the winter of 2017/2018.
 How to calculate border trade constraints? 63

Section of this chapter, topic and relevant Relevant article

regulation
Regulation (EU) 2019/943 adds more tasks for Recital 55 of Regulation (EU) 2019/943 states that ‘The tasks of regional
RSCs, renamed as regional coordination centres. coordination centres should cover the tasks carried out by [RSCs pursuant
to SO GL] as well as additional system operation, market operation and risk
preparedness tasks. The tasks carried out by regional coordination centres
should not include real-time operation of the electricity system.’ In Art. 37(1),
all 16 tasks are listed. Annex I of the same regulation provides additional
information.
The CACM GL includes a methodology to Art. 15 requires CCRs to be determined. A CCR is a geographical area in
define capacity calculation regions (CCRs). which coordinated capacity calculation is applied. Each CCR comprises
a set of bidding zone borders. Finally, ACER (2016a) approved the revised
proposal of all the TSOs, under the condition of merging central western
Europe and central eastern Europe into one Core region.
The CACM GL asks for the development of Art. 20(2) states that ‘no later than 10 months after the approval of the
a coordinated capacity calculation methodology proposal for a capacity calculation region in accordance with Article
for each CCR. 15(1), all TSOs in each capacity calculation region shall submit a proposal
for a common coordinated capacity calculation methodology within the
respective region.’ A methodology for both day-ahead and intraday capacity
calculation is to be proposed. Delayed submission was permitted for certain
CCRs (CACM GL, Art. 20(3–4)). ACER and CEER (2019) report that as of
June 2019 the relevant NRAs approved capacity calculation methodologies in
all CCRs except Italy North.
TSOs remain responsible for maintaining Art. 42 of Regulation (EU) 2019/943 describes the adoption and review
operational security. They can decide to not of coordinated actions and recommendations. This article states that TSOs
implement a coordinated action proposed by shall implement the coordinated actions issued by RCCs except where the
the RCC. They then need to report the detailed implementation would result in a violation of the operational security limits
reason to the RCC and the TSOs of the region. defined by each TSO in accordance with the SO GL. It adds that RCCs may
This will be further investigated. The RCC may propose a different set of coordinated actions in the case that TSOs decide not
then propose a different set of actions. to implement their recommendations.
Section 3.4
The CACM and FCA GL each include Art. 16 of the CACM specifies the GLDPM methodology. Art. 17 of the
a generation and load data provision FCA GL states that the GLDPM developed under the FCA GL shall take into
methodology (GLDPM). account and complement the GLDPM provided in Art. 16 of the CACM GL.
The GLDPM sets out the generation and load Art. 16(2) of the CACM GL states that the proposal for the GLDPM shall
data that TSOs can request from grid users. This specify which generation units and loads are required to provide information
applies mainly to grid users connected at voltage to their respective TSOs for the purpose of capacity calculation. Details on
levels of 220 kV or above. It can also apply to which grid users have to provide what data can be found in the approved
voltage levels below 220 kV if they are used versions of the respective GLDPMs, e.g. All TSOs (2016).
in regional operational security analysis of the
capacity calculation timeframe concerned.
64 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant article

regulation
TSOs can only use the GLDPM to request data Art. 3(1) of the GLDPM approved according to the CACM GL (All TSOs
if that data is not already available to them (2016)) states that ‘Each TSO shall have the right but not the obligation to
through other national or EU legislation (e.g. the obtain these data from the owner of the corresponding network element or the
ENTSO-E Transparency Platform), which can party responsible for providing the information, as the case may be, provided
help to avoid double reporting. that all of the following conditions are met ...’ One of the conditions is that
the data are not already available to the TSO through other national or EU
legislation.
According to the CACM GL, TSOs can only use Art. 16(5) specifies that the information received under the GLDPM shall
the information received under the GLDPM for only be used for capacity calculation purposes.
the purpose of capacity calculation.
The three network codes (CACM GL, FCA GL CACM GL Art. 2(1) defines an IGM as ‘a data set describing power system
and SO GL) introduce individual grid models characteristics (generation, load and grid topology) and related rules to
(IGMs). change these characteristics during capacity calculation, prepared by the
responsible TSOs, to be merged with other individual grid model components
in order to create the common grid model.’CACM GL Art. 19, FCA GL Art.
20 and SO GL Arts. 66 and 70 provide more information about the respective
IGMs.
RCCs merge the IGMs into common grid models According to the CEP, more specifically Regulation (EU) 2019/943, RCCs
(CGMs). take over the tasks of RSCs. Art. 77(3) of the SO GL states that one of the
tasks of the RSCs is to build a common grid model (CGM). The CGM is
defined in the CACM GL as ‘a Union-wide data set agreed between various
TSOs describing the main characteristics of the power system (generation,
loads and grid topology) and rules for changing these characteristics during
the capacity calculation process.’
The three network codes (CACM GL, FCA GL Art. 17 of the CACM GL describes the CGM methodology. Point b in the
and SO GL) describe the IGM merging process second paragraph specifies that this methodology shall contain a description
in the different CGMs. The difference between of the process for merging IGMs to form the CGM. Art. 18 in the FCA GL
these different models lies in the timeframe in introduces the CGM methodology for long-term capacity calculation. Art.
which they are calculated and used. 67 in the SO GL describes the year-ahead CGMs and Art. 70 in the same
regulation describes the methodology for building day-ahead and intraday
CGMs. The SO GL refers to the use of CGMs for various system operation
processes.
It is expected that the RCCs will take turns to Art. 28 of the CACM GL describes the creation of the common grid model.
merge the IGMs into CGMs. Art. 28(2) states that ‘each TSO shall deliver to the TSOs responsible for
merging the individual grid models into a common grid model the most
reliable set of estimations practicable for each individual grid model.’ Art.
77(3) of the SO GL states that one of the tasks of the RSC is to build the
CGM. The CGM is pan-European while the capacity calculation is done for
the CCRs. Therefore, it is expected that the RCCs will rotationally take the
role of the ‘merging agent’.
ENTSO-E stores the IGMs and the CGMs The SO GL describes the general provisions for the OPDE in Art. 114.
in a portal that is called the pan-European
Operational Planning Data Environment (OPDE)
 How to calculate border trade constraints? 65

Section of this chapter, topic and relevant Relevant article

regulation
Following the FCA GL, the CCRs can use the Long-term capacity calculation may be based on the CGM. Alternatively,
CGMs for their capacity calculations in the CCRs may opt to use the statistical approach pursuant to Art. 10(4.b) of the
timeframes before day-ahead, but they can also FCA GL.
opt for a statistical approach.
The SO GL requires a key organizational Art. 40(6) states that ‘all TSOs shall jointly agree on key organisational
requirements, roles and responsibilities requirements, roles and responsibilities in relation to data exchange. Those
(KORRR) methodology. The approved version organisational requirements, roles and responsibilities shall take into account
of this methodology will be decided at the and complement where necessary the operational conditions of the [GLDPM
national level whether distribution-connected developed in accordance with Article 16 of the CACM GL].’ In December
significant grid users (SGUs) have to provide 2018, the revised KORRR methodology proposal by All TSOs (2018) was
their data directly to the TSO or through the approved by all NRAs. Art. 3(3) in this approved methodology states that
connecting DSOs. ‘Subject to approval by the competent regulatory authority or by the entity
designated by the Member State and according to Article 40 of the SO GL,
it shall be determined at a national level whether distribution-connected
SGUs in their TSO’s control area shall provide the structural, scheduled
and real-time data to the TSO directly or through their connecting DSOs or
to both. The decision on the data exchange model may be independent for
each type of information and SGU, if required. When the data is provided
to the DSO, the DSO shall provide the required data to the TSO with
a data granularity necessary to comply with the requirements of the SO GL
provisions.’Art. 2 of the SO GL gives more detail on which SGUs the SO GL
applies to.
Section 3.5
The CACM GL introduced a bidding zone Art. 32 states who can launch a bidding zone review process and how this
review process. process is performed by reviewing existing bidding zone configurations. The
first bidding zone review (ENTSO-E 2018b) was initiated by a letter from
ACER dated 21 December 2016 specifying central Europe as the relevant
region.
Following the CACM GL, the three main criteria The three criteria for reviewing bidding zone configurations are listed and
considered are the stability and robustness of elaborated upon in Art. 33.
the bidding zones, network security and overall
market efficiency.
Regulation (EU) 2019/943 introduces an Regulation (EU) 2019/943 Art. 14 describes the bidding review process.
improved bidding zone review process.
The implementation of this new process has Art. 14(5) in Regulation (EU) 2019/943 states that ‘By 5 October 2019
already started with the submission of a bidding all relevant transmission system operators shall submit a proposal for the
zone review methodology and alternative methodology and assumptions that are to be used in the bidding zone review
configurations by TSOs. process and for the alternative bidding zone configurations to be considered
to the relevant regulatory authorities for approval.’
66 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant article

regulation
According to Regulation (EU) 2019/943, if Art. 14(8) states that ‘For those Member States that have opted to amend
Member States that have opted to change the the bidding zone configuration …, the relevant Member States shall reach
existing bidding zone configuration cannot a unanimous decision within six months of the notification ... Other Member
reach a unanimous decision, as a last resort the States may submit comments to the relevant Member States, which should take
European Commission shall adopt a decision account of those comments when reaching their decision. The decision shall
to amend or maintain the bidding zone be reasoned and shall be notified to the Commission and ACER. In the event
configuration. that the relevant Member States fail to reach a unanimous decision within
those six months, they shall immediately notify the Commission thereof. As
a measure of last resort, the Commission after consulting ACER shall adopt
a decision whether to amend or maintain the bidding zone configuration
in and between those Member States by six months after receipt of such
a notification.’
Member States have to comply with Art. 16(8) states that ‘Transmission system operators shall not limit
a transmission capacity availability threshold. the volume of interconnection capacity to be made available to market
Regulation (EU) 2019/943 states that participants as a means of solving congestion inside their own bidding
the available transmission capacity of all zone or as a means of managing flows resulting from transactions internal
transmission elements possibly limiting the to bidding zones.’ Art. 16(8) continues that the following minimum levels
amount of electricity that can be exchanged of available capacity for cross-zonal trade that have to be reached (at any
between different bidding zones should be at market time unit) are (a) for borders using a coordinated net transmission
least 70 per cent of a benchmark value. capacity approach, the minimum capacity shall be 70 per cent of the
transmission capacity respecting operational security limits after deduction
of contingencies, as determined in accordance with the CACM GL; (b)
for borders using a flow-based approach, the minimum capacity shall be
a margin set in the capacity calculation process as available for flows induced
by cross-zonal exchange. The margin shall be 70 per cent of the capacity
respecting operational security limits of internal and cross-zonal critical
network elements, taking into account contingencies, as determined in
accordance with the CACM. Lastly, Art. 16(8) adds that a total amount of 30
per cent can be used for reliability margins, loop flows and internal flows on
each critical network element.
There are three exceptions to the 70 per cent rule: (a) where structural
congestion has been identified, in the case that Member States opt for
a multinational action plan, the linear trajectory according to Art. 15(2) has
to be reached. Compliance with the 70 per cent rule shall be reached by 31
December 2025; (b) Art. 16(9) states that at the request of TSOs in a CCR,
the relevant regulatory authorities may grant a derogation from the 70 per
cent rule on foreseeable grounds where necessary for maintaining operational
security. This derogation shall be granted for no more than one year at a time
or, provided that the extent of the derogation decreases significantly after
the first year, up to a maximum of two years; (c) Art. 16(3) adds that where
RCCs conclude that the available remedial actions in the capacity calculation
region or between capacity calculation regions are not sufficient to reach
the minimum capacities provided (or the linear trajectory) while respecting
operational security limits, they may, as a measure of last resort, set out
coordinated actions reducing the cross-zonal capacities accordingly.
 How to calculate border trade constraints? 67

Section of this chapter, topic and relevant Relevant article

regulation
A process has been set in motion to assess the Note that the 70 per cent minimum capacity available for trade in Regulation
extent to which different countries already (EU) 2019/943 applies from January 2020, the date of entry into application
comply with this threshold and how they intend of Art. 71(2). In addition, Art. 15(2) says that if a Member State opts for
to comply with the threshold by 2025. an action plan, the minimum capacity, as described in Art. 16(8), shall be
reached by 31 December 2025.
Art. 16(3) states that RCCs shall submit a report to the relevant regulatory
authorities and to ACER on any reduction of capacity or deviation from
coordinated actions due to operational security threats. The RCCs shall assess
the incidences and make recommendations, if necessary, on how to avoid
such deviations in the future.
ACER and CEER (2019) are also already monitoring compliance with the 70
per cent rule in their market monitoring report.
Countries need to choose between the two DG Art. 14(7) states that where structural congestion has been identified, the
COMP cases. They can follow the Swedish or Member State concerned shall, in cooperation with its TSOs, decide to
the German approach to stop discrimination establish national or multinational action plans, or to review and amend its
between cross-border and national trade. bidding zone configuration.
Structural congestion is defined in Art. 2(6) as ‘congestion in the transmission
system that is capable of being unambiguously defined, is predictable, is
geographically stable over time, and frequently reoccurs under normal
electricity system conditions.’ The identification of long-term structural
congestion can be done in three ways: using ENTSO-E’s report on structural
congestion (Art. 14(2)); using TSO assessments; or the bidding zone review
itself (Art. 14(7)).
According to Art. 16(8), in the case that a Member State fails to respect the 70
per cent rule while not opting for a multinational action plan or if a Member
State has opted for a multinational action plan but does not respect the linear
trajectory, a bidding zone reconfiguration can result (Arts. 15(7) and 15(5)).
4. Who pays for the network when trade is
international?
Leonardo Meeus with Tim Schittekatte

In this chapter, we answer four questions. First, who pays for the network? Second, why did
national network tariffs start to be harmonized? Third, why was there a move away from
transit charges? And fourth, how to share network investment costs between countries?

4.1 WHO PAYS FOR THE NETWORK?

Electricity networks are important infrastructure like roads, railways and fibre networks. Such
important infrastructure is typically paid for through a combination of general taxation and
user tariffs. However, electricity networks in Europe are mostly paid for by network users
through so-called network tariffs. In what follows, we introduce the two types of network
tariffs: connection charges and access charges.1
First, connection charges. As the name indicates, these charges are for connection to the grid
at the time of connecting and are typically one-off payments that in some cases can be spread
over time. They are often labelled ‘super-shallow’, ‘shallow’ or ‘deep’ connection charges.
Super-shallow means a very cheap or free connection. Shallow means you pay for the cable
and other necessary equipment to connect you to the electricity network’s local feeder. You
are the only user of that piece of the network so the cost can easily be attributed to you. Deep
means that you also pay for network reinforcements that might be needed deeper inside the
network to accommodate the new connection at your location. Deep charges try to influence
the location of new connections by signalling where the network can still host additional
connections without reinforcement and where reinforcement would be needed. Shallow
connection charges are the dominant model in Europe, but there are also a few countries with
deep charges (five countries in 2018). Connection charges can also vary within a country for
different voltage levels and/or different grid users. Consumers and producers are not necessar-
ily treated in the same way.
Second, access charges. These charges are paid on a monthly or bi-annual basis. Network
access charges are usually divided between transmission access charges and distribution
access charges. Grid users contribute to the network cost of the voltage level that they are
connected to and the voltage levels above them. This means that grid users connected to the
highest transmission-level voltage only pay transmission charges for that level, while house-
holds connected to the lowest distribution-level voltage pay for all the network levels above
them through both transmission and distribution access charges. The origin of this cascading
principle is that electricity flows from the highest voltage all the way down to the lowest.
Transmission access charges are also used to recover the costs of some of the services that are
68
 Who pays for the network when trade is international? 69

provided by transmission system operators (TSOs). The European Network of Transmission

System Operators for Electricity (ENTSO-E) reports that on average in Europe in 2018, 59 per
cent of transmission access charges covered infrastructure costs, 31 per cent system services
and 10 per cent losses.
TSOs also collect congestion revenue, which can cover part of their costs and thus can
reduce the need to collect money through network tariffs. This revenue results from explicit
or implicit auctioning of transmission rights, as was discussed in Chapter 2. According to
ENTSO-E data, on average €2 billion of congestion revenue a year was collected in the period
2011–2015. In Annex 4A.1, we provide an example of how this congestion revenue is calcu-
lated under implicit auctioning of transmission rights.

4.2 WHY DID NATIONAL NETWORK TARIFFS START TO BE

HARMONIZED?

In this section, we first discuss the level of harmonization of transmission network tariffs and
then introduce two open issues.

4.2.1 The Harmonization of Transmission Network Tariffs

For the moment, network tariffs have only been harmonized at the transmission level, and only
access charges, not connection charges, are harmonized. Moreover, only the access charges
that apply to generators have to a certain extent been harmonized (the so-called G-component),
not those that apply to consumers (the C-component). Not all countries have a G-component,
but those that do (15 countries in 2018) have to comply with the caps listed in Table 4.1. These
caps apply to the total amount of money that is collected from producers divided by their
total output per year. Connection charges, charges related to system services and losses are
excluded from this calculation. These caps were introduced in Regulation (EU) No 838/2010.
The development of guidelines around the harmonization of transmission tariffs was first
described in Regulation (EC) No 1228/2003, which was part of the Second Energy Package.
The main argument for a G-component in transmission access charges is that it ensures
that generators consider transportation costs in their decisions. Connection charges can guide
generators to locate where it would be cheaper to transport what they produce. Access charges
can also guide decisions to produce or consume in the period after the connection. When the
network is in a critical condition, generators can be given a signal not to produce. The imple-
mentation of a geographically differentiated G-component is limited to the UK, Ireland and
Sweden. The argument against a G-component is that generators in that country are at a com-
petitive disadvantage with respect to generators that access the European electricity market via
a country that does not have such a G-component. It is this distortion of the level playing field
that motivated the introduction of the above-mentioned caps. Moreover, locational signals can
be more effectively provided through variations in energy prices over well-defined zones (or
nodes) as described in Chapter 3.
A similar logic can be applied to distribution network tariffs. Indeed, producers increasingly
access the European electricity market through distribution grids. In our research, we made
a first attempt to calculate the spillover effects caused by distribution tariffs and concluded that
they can be significant.2 During the negotiations on the Clean Energy Package, the European
70 Evolution of electricity markets in Europe

Table 4.1 Maximum annual average transmission charges paid by producers

according to Regulation (EU) No 838/2010

Maximum annual average transmission charge paid by

producers
Great Britain, Ireland and Northern Ireland €2.5/MWh
Romania €2/MWh
Denmark, Finland and Sweden €1.2/MWh
All other EU countries €0.5/MWh

Commission proposed harmonizing distribution network tariffs through the EU network code
process, but this proposal was dropped due to strong opposition on the part of stakeholders and
national governments.

4.2.2 Network Tariffs – Open Issues

In this subsection we introduce two open issues related to network tariffs: the treatment of
energy storage and the implementation of the cost-reflectivity principle.3
First, the treatment of energy storage. In some countries, energy storage is charged like gen-
eration and in other countries like consumption. There are also countries where energy storage
is charged for consumption and for generation, and countries that have started to treat storage
separately when it comes to network access charges. For instance, Belgium has exempted
newly connected energy storage assets from transmission charges for ten years. Charges that
apply to generators to access the European electricity market to create a level playing field
have been harmonized so the same logic could be applied to energy storage. We expect energy
storage to play an important role in the electricity system of the future, so we can expect that
debate on this issue will continue.
Second, the implementation of the cost-reflectivity principle. Regulation (EU) 2019/943
mandates ACER to come up with best practice reports on transmission and distribution network
tariffs. The regulation also states that these tariffs should become more cost-reflective, which
can include locational and time-varying signals. Historically, distribution access charges were
simple rather than cost-reflective. Consumers paid in proportion to the volume of their annual
consumption in euro/kWh. With the emergence of rooftop photovoltaic (PV) panels, electric
vehicles, heat pumps and home batteries, simple tariffs increasingly give distorted signals.
The best-known example is what happened with the large-scale introduction of rooftop PVs.
Wealthy households used to consume more so they contributed more to distribution network
costs. Today, volumetric tariffs achieve the opposite. Wealthy households can more easily
invest in rooftop PVs. In some hours of the day, these households inject the output of their PV
panels into the network, and in other hours they withdraw that energy back from the network.
If we only measure and charge on the basis of the net effect, they no longer pay for the network
they continue to use at the expense of other users who cannot afford PV panels.
Cost-reflective tariffs mean you pay for the costs you cause. This is easier said than done,
and we also do not want tariffs to become too complicated. Cost-reflectivity can become
politically sensitive if it implies different charges for urban and rural areas, or if it implies
revisiting the allocation of network costs between households and industry. Following the
 Who pays for the network when trade is international? 71

cascading principle, households connected to distribution networks contribute to transmission

network costs, while large industrial consumers connected to transmission networks do not
contribute to distribution network costs, which could be challenged in an electricity system
that is increasingly decentralizing.

4.3 WHY WAS THERE A MOVE AWAY FROM TRANSIT

CHARGES?

In this section we first introduce transit charges and then discuss the inter-TSO compensation
(ITC) scheme that replaced them.4
First, transit charges. Who pays for the network becomes more complicated when you
introduce international trade. As discussed in the previous sections, national network tariffs
mean charging producers and consumers connected to the transmission and distribution
network when they connect or when they access the grid. However, what should be done with
traders that use the network to transfer energy from one national network to another, possibly
also crossing the networks of transit countries? Most countries introduced a so-called transit
charge to apply to cross-border transactions. This meant that traders not only needed to buy
transmission rights to be able to trade across borders, but they also had an extra transaction
cost for trading across the border.
However, cross-border transactions only cause network costs to the extent that they cause
flows. There are many more transactions than physical flows, and transactions can go both
ways across a certain border so that they cancel out and do not cause any flow or cost. It is
also possible to have physical flows across a network even if there are no transactions. Traders
can transact via another path around a country while the flow will still go through its network.
Transit charges also allowed countries to tax international trade in order to reduce their national
network tariffs. Transit charges were an obstacle to trade as they were increasing transaction
costs rather than charging the users of the grid. Therefore, they were finally abolished in 2014.
The process that led to the abolishment of transit charges is interesting, as it played out over
many years in the Florence Forum. Stakeholders started to meet informally in Florence and the
gathering was then institutionalized by the European Commission. In 2002, the 8th Florence
Forum agreed to abolish all existing transit charges while still allowing capped export charges.
Initially the cap on export charges was set at €1/MWh, but was then reduced to €0.5/MWh.
Finally, this led to their abolishment in 2004, long before roaming was abolished for mobile
phones, which received much more attention.
Second, the inter-TSO compensation (ITC) scheme. In 2002, eight European TSOs signed
a first ITC agreement. Through their associations (ETSO and later ENTSO-E), TSOs con-
tinued to develop and expand this mechanism. Under this scheme, TSOs from countries that
cause international flows by importing or exporting contribute to a fund and TSOs from coun-
tries that host international flows receive money from that fund. TSOs that pay into the fund
collect the money by increasing their national network tariffs and those that receive money
from the fund use it to lower their national network tariffs. It is a zero-sum game that pays for
the costs caused by transit flows and avoids the problems with transit charges on transactions.
Even though the principle of the ITC scheme is clear, its implementation has proven to be
challenging. To determine the size of the fund, TSOs and their national regulatory authorities
(NRAs) need to agree on the level of costs, and there are different cost-reporting practices. The
72 Evolution of electricity markets in Europe

definition of a transit has also been challenged. There can be transit flows that cross a country,
but there can also be transit flows that enter a country to run along the border and exit again.
Both would count equally as transits, but they do not cause the same level of costs. It therefore
took a long time to agree on a definitive ITC methodology. Note that transit flow is a concept
that is also used in the debate on flow-based market coupling (see Annex 3A.1 and Annex
3A.2 in Chapter 3), but the definition is not the same as in the debate on the ITC methodology.
As with G-charges in the previous section, the ITC mechanism was already part of Regulation
(EC) No 1228/2003 and it was further formalized in Regulation (EU) No 838/2010.
In other words, the ITC scheme is far from perfect but it did help to abolish transit charges,
which were worse. Sensitivities related to transit charges and the ITC scheme dominated the
debate for a long time, but gradually faded. In 2017, the ITC fund amounted to €259.3 million.
This consisted of €100 million to compensate for infrastructure costs and €159.3 million to
compensate for losses caused by transits. In comparison with the schemes we will discuss in
the next section and that have received most of the attention in recent years, the money that
circulates in the ITC mechanism is relatively small and not used to provide investment signals
but instead to allocate sunk investments costs.

4.4 HOW TO SHARE NETWORK INVESTMENT COSTS

BETWEEN COUNTRIES?

In this section we first introduce the process of prioritizing projects of common interest (PCI)
in Europe and then discuss so-called cross-border cost allocation (CBCA) agreements for
these projects, which determine how network investment costs are charged across countries.
First, projects of common interest (PCIs). Following Regulation (EC) No 714/2009, TSOs
make national transmission investment plans and they are consolidated into the so-called
Ten-Year Network Development Plan (TYNDP).5 The TYNDP is one of ENTSO-E’s tasks,
and in 2018 it argued that €114 billion of investment in power lines would be needed by
2030. Regulation (EU) No 347/2013, better known as the Trans-European Energy Networks
(TEN-E) Regulation, then introduced a process in which some of these projects could become
PCIs. The list of PCIs also includes third-party projects. Third-party projects can be merchant
projects that have followed an exemption process or regulated projects from countries that
allow third parties to compete with TSOs to develop such projects. The first PCI list was pub-
lished in 2013. It is updated every two years and contains a selection of electricity transmission
lines, gas and oil pipelines, gas and electricity storage projects, electricity smart grids, and
carbon dioxide transport infrastructure. The total investment cost of the electricity projects on
the 2017 PCI list is €49.3 billion – almost half of the TYNDP investment in 2018.
To obtain PCI status, projects need to conduct a cost–benefit analysis (CBA) following
a methodology designed by ENTSO-E with supervision by ACER. They can then be ranked
on the basis of their CBA and be selected in a process that is set out in the TEN-E Regulation.
Projects with PCI status have advantages, such as accelerated permitting procedures, dedicated
financial incentives from the NRAs involved and access to the Connecting Europe Facility
(CEF). In the period 2014–2020, a total of €5.35 billion in EU funding was allocated to PCI
projects via the CEF.
Second, cross-border cost allocation (CBCA) agreements. Countries used to agree on
cross-border investments on the assumption that they would each pay for assets in their ter-
 Who pays for the network when trade is international? 73

ritories. If they both benefited enough to justify these costs, they would agree to go forward
with the investment. If one of them had doubts, the project would be cancelled or delayed.
The country that was more convinced about the project then had an incentive to compensate
the neighbour. In Box 4.1 we illustrate two projects where countries started to do this on
a voluntary basis.
With the CBCA process, the TEN-E Regulation mandated ACER to act as a mediator in
these kinds of cases. TSOs make a CBCA proposal for PCIs, and if the NRAs cannot agree
ACER can intervene. This has happened twice so far, namely the Gas Interconnection Poland–
Lithuania (GILP) in August 2014 and the Lithuanian part of the Electricity Interconnection
between Lithuania and Poland (LitPol Link) in April 2015. ACER decided that a case where
the default cost allocation clearly results in a net loser, that is, a country that benefits less than
the costs it is expected to incur by investing in the assets in its territory, is not a justification
for applying for CEF funding. Instead, the other beneficiaries are first asked to compensate the
net loser for its loss, and other beneficiaries can even be third countries which do not host any
of the assets in their territory but clearly benefit from the projects. Only in the GILP case did
ACER allocate costs beyond the hosting countries, according to its principle of compensating
net losers. In the case of the LitPol Link, ACER decided not to allocate costs to non-hosting
countries because there was no net loser.
Our research provides a more detailed discussion of these cases because we had the pleasure
of advising ACER in this process. Note that the allocation of CEF funding by the European
Commission remains controversial. The CBA and CBCA process tried to reduce the politics,
but the impression is that this has only been partly accomplished.

BOX 4.1 TWO EXAMPLES OF INNOVATIVE CBCA PRACTICES

The Norway–Sweden case

Norway is divided into five bidding zones because it has structural congestion within the
country. In dry years, the energy supply within each bidding zone can be very tight. This is
the case in mid-Norway, but the situation worsened in 2005 because of new industrial con-
sumption. The situation became critical in the following dry year. The easiest and quickest
solution for Norway was to increase the interconnection capacity between mid-Norway
and Sweden (line A on the left-hand side of Figure 4.1). This is a 100-km-long 420 kV AC
line between Nea and Järpströmmen, which was commissioned in 2009. Seventy-five per
cent of the assets are in Swedish territory. Line B, the line in Norwegian territory (on the
left-hand side of Figure 4.1) was expected to increase the available capacity on line A from
200 to 750 MW so that Sweden would also benefit from the cross-border exchange, but line
B was less advanced. Norway therefore agreed to compensate Sweden for line A until line
B was ready.
The compensation incentivized Sweden to speed up the development of line A, and also
incentivized Norway to speed up the development of line B. The TSOs involved entered
into a formal contract, which was approved by their NRAs.
74 Evolution of electricity markets in Europe

Figure 4.1 The Norway–Sweden case (left) and the Italy–Greece case (right)

The Italy–Greece case

In 2002, a 500 MW submarine high voltage direct current (HVDC) link between Greece
and Italy (GRITA) was commissioned (see the right-hand side of Figure 4.1). Italy paid for
and owns 75 per cent of the project and Greece the remaining 25 per cent. The project also
received an EU grant, so this CBCA agreement only applied to part of the project costs.
The project allows Italy to import cheaper electricity from eastern European countries, like
Albania and Turkey, via Greece.
This CBCA showcases innovation, with parties deviating from the common 50:50 cost
sharing for an interconnector. It is a typical case in which a transit country (Greece) is com-
pensated to jointly develop a project with its neighbour (Italy).

4.5 CONCLUSION

In this chapter on network cost allocation when trade is international, we have answered four
questions.
First, who pays for the network? The common practice today for electricity networks is that
they are mostly paid for by the network users. There are two types of network tariffs: connec-
tion charges and network access charges.
Second, why did national network tariffs start to be harmonized? Transmission network
tariffs were somewhat harmonized by Regulation (EU) No 838/2010, which introduced a cap
on access charges for generators. Currently, how to apply network tariffs to storage and how to
implement the cost-reflectivity principle in distribution network tariffs are open issues.
Third, why was there a move away from transit charges? Transactions do not cause network
costs – only flows do; so transit charges were increasing transaction costs rather than charging
the users of the grid. Under the inter-TSO compensation scheme, TSOs pay in proportion to
the international flows they cause and receive money in proportion to the international flows
they host.
Fourth, how to share network investment costs between countries? Each country used to
pay for the assets on its territory even if it was not the main beneficiary. Following the TEN-E
Regulation, projects can receive project of common interest status. PCIs have several advan-
 Who pays for the network when trade is international? 75

tages, including the cross-border cost allocation (CBCA) process in which ACER can mediate
in projects to reach a better cost allocation that can help unblock them.

NOTES
1. In ENTSO-E (2018) there is an overview of transmission tariffs in Europe. This report is updated
every year and the facts and figures on transmission tariffs quoted in this chapter come from this
report. In Ruester et al. (2012) we provided an academic discussion on transmission tariffs and the
possible role of the EU.
2. In Govaerts et al. (2019) we provided a first analysis of the spillover effects of distribution grid
tariffs in the internal electricity market. A simplified numerical example is used to give insight into
the order of magnitude of the spillovers and the main sensitivities that drive these effects.
3. In Schittekatte et al. (2018) and Schittekatte and Meeus (2020) we discuss why future-proof
distribution tariffs need to anticipate the behaviour of households in response to price signals. We
also discuss the distributional effects of tariffs by analysing their impact on active versus passive
customers. Our colleagues from Comillas and MIT have also done a very interesting analysis on this
topic in the ‘Utility of the Future’ study by the MIT Energy Initiative (2016). In CREG (2018) and
Ofgem (2019), the treatment of energy storage for transmission network tariffs is discussed in detail.
4. In ACER (2018b) and ENTSO-E (2019a) there is detailed information on the ITC mechanism.
These reports are updated every year with the latest statistics. For more background, refer to
ERGEG (2006). Also interesting is academic work on the topic. Olmos and Pérez-Arriaga (2007),
Pérez-Arriaga et al. (2002) and Daxhelet and Smeers (2007) analyse the different options that have
been considered to implement the ITC scheme. In the conclusions of the 8th and 10th Florence Fora
published by the European Commission (2002, 2003), the gradual abolishment of transit charges
can be observed. The prominence of this issue on the agenda over many years can also be seen from
the conclusions of the earlier Florence Fora.
5. The TYNDP is described in ENTSO-E (2019b). The TYNDP is updated every two years. ACER
(2018a) reports on the two-yearly PCI list. European Commission (2018) is an example of a press
release related to the Connecting Europe Facility (CEF) funding scheme. In Keyaerts et al. (2016)
and Meeus et al. (2013) we gave our recommendations for the development of the cost–benefit
analysis (CBA) methodology that is used for PCIs in Europe. We also supported ACER in its
approach to cross-border cost allocation (CBCA) (ACER, 2015b). In Meeus and He (2014) we
made recommendations to ACER when it was given the CBCA mandate. We argued that ACER
could only intervene in certain cases to encourage stakeholders to come forward with innovative
approaches. The examples shown in Box 4.1 originate from that policy brief. One year later, in
Meeus and Keyaerts (2015) we also evaluated the first cases that were treated by the ACER. In
Keyaerts and Meeus (2017) we provide case studies on countries that decided to introduce dedicated
financial incentives for priority projects. In Bhagwat et al. (2019) we updated our CBA and CBCA
analysis with a focus on the implications for meshed offshore transmission networks to support
the development of wind power. In Schittekatte et al. (2020) we provide recommendations on how
to revise the TEN-E Regulation to be aligned with the objectives of the European Green Deal.
Furthermore, for a study on how the incentives of national TSOs are not necessarily aligned with
the overall European welfare when deciding about building an interconnector, see the work of Matti
Supponen (2011). Finally, for deeper discussions about the implications of the Trans-European
Energy Networks (TEN-E) Regulation and its importance, see the book edited by Jean Arnold
Vinois (2014). For more details about the two ACER decisions, please consult ACER (2014,
2015a).

REFERENCES
ACER (2014), ‘Decision No 01/2014 of the Agency for the Cooperation of Energy Regulators of
11 August 2014 on the Investment Request including Cross-Border Cost Allocation for Gas
Interconnection Poland–Lithuania. Project of Common Interest No 8.5’.
76 Evolution of electricity markets in Europe

ACER (2015a), ‘Decision No 02/2015 of the Agency for the Cooperation of Energy Regulators of 16
April 2015 on the Investment Request Including Cross-Border Cost Allocation for the Lithuanian Part
of the Interconnection between Alytus (LT) and the Lithuanian–Polish Border. Project of Common
Interest No 4.5.1’.
ACER (2015b), ‘Recommendation No 05/2015 of the Agency for the Cooperation of Energy Regulators
of 18 December 2015 on Good Practices for the Treatment of Investment Requests, including CBCA
Requests, for Electricity and Gas PCIs’.
ACER (2018a), ‘Consolidated Report on the Progress of Electricity and Gas Projects of Common
Interest for the Year 2017’, published on 10 July 2018.
ACER (2018b), ‘ITC Monitoring Report 2018’, Report to the European Commission on the
Implementation of the ITC Mechanism in 2017, published in October 2018.
Bhagwat, P., T. Schittekatte, L. Lind, N. Keyaerts and L. Meeus (2019), ‘D7.4 Economic Framework for
a Meshed Offshore Grid’, Final Report PROMOTioN.
CREG (2018), ‘Besluit tot vaststelling van de tariefmethodologie voor het elektriciteits-transmissienet
en voor de elektriciteitsnetten met een transmissiefunctie voor de regulatoire periode 2020–2023’
[‘Conclusion on the Application of Tariff Methodologies for Electricity Transmission Systems and
Electricity Transmission Functions for the Regulatory Period 2020–2023’], Beslissing(Z)1109/10,
published on 28 June 2019.
Daxhelet, O. and Y. Smeers (2007), ‘The EU regulation on cross-border trade of electricity: A two-stage
equilibrium model’, European Journal of Operational Research, 181 (3), 1396–412.
ENTSO-E (2018), ‘Overview of Transmission Tariffs in Europe: Synthesis 2018’, Annual Report,
published in May 2018.
ENTSO-E (2019a), ‘ITC Transit Losses Data Report 2018’, Annual Report, published on 24 September
2019.
ENTSO-E (2019b), ‘TYNDP 2018 – Executive Summary’, Connecting Europe: Electricity
2025–2030–2040. Final Version after Consultation and ACER Opinion – October 2019.
ERGEG (2006), ‘Cover note to ERGEG draft proposal on Guidelines on Inter TSO Compensation’, Ref:
E06-CBT-09-08a, 10 April 2006.
European Commission (2002), ‘European Electricity Regulatory Forum: A significant step forward in
completing the internal electricity market’, MEMO/02/37. Brussels, 27 February 2002, accessed at
https://​ec​.europa​.eu/​commission/​presscorner/​detail/​en/​MEMO​_02​_37.
European Commission (2003), ‘Florence Forum gets rid of all cross border electricity charges’,
MEMO/03/153. Brussels, 18 July 2003, accessed at https://​ec​.europa​.eu/​commission/​presscorner/​
detail/​en/​MEMO​_03​_153.
European Commission (2018), ‘More growth and jobs: EU invests €873 million in clean energy infra-
structure’, press release of 25 January 2018, accessed at https://​ec​.europa​.eu/​commission/​presscorner/​
detail/​en/​IP​_18​_383.
Govaerts, N., K. Bruninx, H. Le Cadre, L. Meeus and E. Delarue (2019), ‘Spillover effects of distribution
grid tariffs in the internal electricity market: An argument for harmonization?’, Energy Economics, 84,
accessed at https://​doi​.org/​10​.1016/​j​.eneco​.2019​.07​.019.
Keyaerts, N. and L. Meeus (2017), ‘The regulatory experience of Italy and the United States with dedi-
cated incentives for strategic electricity transmission investment’, Utilities Policy, 46, 71–80.
Keyaerts, N., T. Schittekatte and L. Meeus (2016), ‘Standing still is moving backward for the ABC of
the CBA’, FSR Policy Brief 2016/08.
Meeus, L. and X. He (2014), ‘Guidance for project promoters and regulators for the cross-border cost
allocation of projects of common interest’, FSR Policy Brief 2014/02.
Meeus, L. and N. Keyaerts (2015), ‘First series of cross-border cost allocation decisions for projects of
common interest: Main lessons learned’, FSR Policy Brief 2015/01.
Meeus, L., N.-H. von der Fehr, I. Azevedo, X. He, L. Olmos and J.-M. Glachant (2013), ‘Cost Benefit
Analysis in the Context of the Energy Infrastructure Package’, Final Report. THINK Project, Topic
10, published in January 2013.
MIT Energy Initiative (2016), ‘Utility of the Future: An MIT Energy Initiative Response to an Industry
in Transition’.
 Who pays for the network when trade is international? 77

Ofgem (2019), ‘Open letter on implications of charging reform on electricity storage’, Open Letters and
Correspondence, published on 24 January 2019.
Olmos, L. and I. Pérez-Arriaga (2007), ‘Evaluation of three methods proposed for the computation of
inter-TSO payments in the internal electricity market of the European Union’, IEEE Transactions on
Power Systems, 22 (4), 1507–22.
Pérez-Arriaga, I., L. Olmos and F. J. Rubio (2002), ‘Report on Cost Components of Cross Border
Exchanges of Electricity’, Final Report. Prepared for the Directorate-General for Energy and
Transport/European Commission, published in November 2002.
Ruester, S., C. von Hirschhausen, C. Marcantonini, X. He, J. Egerer and J. M. Glachant (2012), ‘EU
Involvement in Electricity and Natural Gas Transmission Grid Tarification’, Final Report. THINK
Project, Topic 6, published in January 2012.
Schittekatte, T. and L. Meeus (2020), ‘Least-cost distribution network tariff design in theory and prac-
tice’, Energy Journal, 41 (5), 97–133.
Schittekatte, T., I. Momber and L. Meeus (2018), ‘Future-proof tariff design: Recovering sunk grid costs
in a world where consumers are pushing back’, Energy Economics, 70, 484–98.
Schittekatte, T., A. Pototschnig, L. Meeus, T. Jamasb and M. Llorca (2020), ‘Making the TEN-E regula-
tion compatible with the Green Deal: Eligibility, selection, and cost allocation for PCIs’, FSR Policy
Brief 2020/27.
Supponen, M. (2011), ‘Influence of National and Company Interests on European Electricity Transmission
Investments’, Aalto University Publication Series Doctoral Dissertations, 77/2011.
Vinois, J. A. (ed.) (2014), EU Energy Law Volume VIII – The Energy Infrastructure Policy of the
European Union, Deventer, Netherlands and Leuven, Belgium: Claeys & Casteels.
78 Evolution of electricity markets in Europe

4A.1 ANNEX: CONGESTION RENT UNDER MARKET

COUPLING: A NUMERICAL EXAMPLE

Suppose that the day-ahead market auction for a certain hour results in a price in zone A of
€50/MWh and a price in zone B of €60/MWh. The satisfied demand in zone A is 100 MWh,
that in zone B is 150 MWh and 50 MW of transmission rights are used to trade between
the two zones in that hour. In this case, electricity flows from the low-price zone (A) to the
high-price zone (B). There is a price difference, which means that the transmission rights have
been fully utilized but they were not enough to create a single price zone. The rights therefore
have a value and congestion rent is collected, as is illustrated in Table 4A.1.

Table 4A.1 A numerical example of congestion rent calculation

Price Demand Generation Demand expense Generation income

150 MWh
Zone A €50/MWh 100 MWh (demand zone A + €5000 €7500
interconnector)
100 MWh
Zone B €60/MWh 150 MWh (demand zone B €9000 €6000
– interconnector)
€14 000 €13 500

The total income for generation over the two zones is €13 500 while the total amount spent by
demand equals €14 000. The difference between the two is the congestion rent of €500, equal-
ling the price differential between the two zones (€10/MWh) multiplied by the capacity of the
line (50 MW). This congestion rent is transferred to the TSO(s) owning the interconnector.
There are strict rules around the use of congestion rent in the EU. Congestion rent should be
used to invest in the network; only if the revenues cannot be used efficiently for investments
can they be used to lower the national transmission network tariffs.
 Who pays for the network when trade is international? 79

4A.2 ANNEX: REGULATORY GUIDE

Table 4A.2 Regulatory guide

Section of this chapter, topic and relevant Relevant article

regulation
Section 4.1
TSOs collect congestion revenues, which can Regulation (EC) No 714/2009 states in Art. 16(6) that any revenue resulting
cover part of their costs, and thus they can from the allocation of interconnection shall be used for guaranteeing the
reduce the need to collect money through actual availability of the allocated capacity and/or maintaining or increasing
network tariffs. interconnection capacities through network investment, in particular in new
interconnectors. The article continues by stating that ‘if the revenues cannot
be efficiently used for these [the two above-mentioned purposes], they may
be used, subject to approval by the regulatory authorities of the Member
States concerned, up to a maximum amount to be decided by those regulatory
authorities, as income to be taken into account by the regulatory authorities
when approving the methodology for calculating network tariffs and/or fixing
network tariffs.’
Section 4.2
Caps on G-charges were introduced in In Art. 8(3) of Regulation (EC) No 1228/2003 it is stated that, where
Regulation (EU) No 838/2010. The development appropriate, the Commission shall adopt and amend guidelines that determine
of guidelines for the harmonization of appropriate rules leading to a progressive harmonization of the underlying
transmission tariffs was first described in principles for the setting of charges applied to producers and consumers
Regulation (EC) No 1228/2003 in the Second (load) under national tariff systems. Regulation (EU) No 838/2010 contains
Energy Package. in its Annex part B ‘Guidelines for a Common Regulatory Approach to
Transmission Charging’ in which the caps on G-charges are specified.
Regulation (EU) 2019/943 and network charges Art. 18(1) states that network charges shall not discriminate either positively
for storage. or negatively against energy storage.
Regulation (EU) 2019/943 mandates ACER Art. 18(9) states that ‘By 5 October 2019 in order to mitigate the risk
to come up with best practice reports on of market fragmentation ACER shall provide a best practice report on
transmission and distribution network tariffs. transmission and distribution tariff methodologies while taking account of
national specificities.’ Furthermore, the article lists several dimensions of
network tariffs that were to be addressed, among which are time and location
differentiation. It also states that ACER shall update the best practice report at
least once every two years. Art. 18(10) adds that ‘Regulatory authorities shall
duly take the best practice report into consideration when fixing or approving
transmission tariffs and distribution tariffs or their methodologies in
accordance with Article 59 of Directive (EU) 2019/944.’Art. 59 of Directive
(EU) 2019/944 lists the duties and powers of the regulatory authorities.
80 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant article

regulation
Regulation (EU) 2019/943 states that these Art. 18(7) states that ‘Distribution tariffs shall be cost-reflective taking into
tariffs should become more cost-reflective, account the use of the distribution network by system users including active
which can include locational and time-of-use customers. Distribution tariffs may contain network connection capacity
signals. elements and may be differentiated based on system users’ consumption
or generation profiles. Where Member States have implemented the
deployment of smart metering systems, regulatory authorities shall consider
time-differentiated network tariffs when fixing or approving transmission
tariffs and distribution tariffs or their methodologies in accordance with
Article 59 of (EU) 2019/944 and, where appropriate, time-differentiated
network tariffs may be introduced to reflect the use of the network, in
a transparent, cost efficient and foreseeable way for the final customer.’
Section 4.3
Transit charges were finally abolished in 2004. Regulation (EC) No 1228/2003, which entered into force on 1 July 2004,
formalized the abolishment of transit charges in Art. 4(5) by stating that ‘there
shall be no specific network charge on individual transactions for declared
transits of electricity.’
As with G-charges in the previous section, the In Art. 8(2) of Regulation (EC) No 1228/2003 it is stated that, where
ITC mechanism was already part of Regulation appropriate, the Commission shall adopt and amend guidelines that shall
(EC) No 1228/2003 and was further formalized specify the details of the ITC mechanism. Regulation (EU) No 838/2010
in Regulation (EU) No 838/2010. contains in its Annex part A ‘Guidelines on the Inter-Transmission System
Operator Compensation Mechanism.’
Section 4.4
Following Regulation (EC) No 714/2009, Art. 8(3.b) states that ENTSO-E shall adopt ‘a non-binding Community-wide
TSOs make national transmission investment ten-year network development plan, including a European generation
plans. These are consolidated into the so-called adequacy outlook, every two years.’
Ten-Year Network Development Plan (TYNDP).
Regulation (EU) No 347/2013, better known as Art. 2 defines a PCI as ‘a project necessary to implement the energy
the Trans-European Energy Networks (TEN-E) infrastructure priority corridors and areas set out in Annex I and which is
Regulation, introduced a process through which part of the Union list of projects of common interest referred to in Article 3.’
some of these projects could become projects of Art. 3 sets out how the Union list of PCIs will be made up. The Commission
common interest (PCIs). is empowered to adopt delegated acts that establish the Union list of PCIs.
The Commission ensures that the Union list is established on the basis of the
regional lists adopted by the decision-making bodies of the Groups. There are
12 regional groups listed in Annex I.
The PCI list also includes third-party projects. The definition of project promoters in Art. 2(6) of Regulation (EU) No
Third-party projects can be merchant projects 347/2013 clarifies that a project promoter can not only be a TSO, DSO or
that have followed an exemption process or other operator but also an investor developing a project of common interest.
regulated projects from countries that allow third
parties to compete with TSOs to develop such
projects.
The PCI list is updated every two years and Art. 3 of Regulation (EU) No 347/2013 states that the Commission shall
contains a selection of electricity transmission ensure that the Union PCI list is established every two years. Annex II of this
lines, gas and oil pipelines, gas and electricity regulation sets out the energy infrastructure categories which are eligible as
storage projects, electricity smart grids, and PCIs.
carbon dioxide transport infrastructure.
 Who pays for the network when trade is international? 81

Section of this chapter, topic and relevant Relevant article

regulation
Regulation (EU) No 347/2013 states that to Annex III.2(1) setting out the process for establishing regional lists states that
obtain PCI status projects need to conduct ‘for projects having reached a sufficient degree of maturity, a project-specific
a cost–benefit analysis (CBA) following cost–benefit analysis in accordance with Articles 21 and 22 based on the
methodology designed by ENTSO-E with methodologies developed by the ENTSO for electricity or the ENTSO for
supervision by ACER. gas pursuant to Article 11’ will be conducted. Art. 4(2) states that all ‘the
potential overall benefits of the project, assessed according to the respective
specific criteria in paragraph 2, outweigh its costs, including in the longer
term.’ Art. 4(2) lists specific criteria that apply to projects of common interest
depending on the specific energy infrastructure categories.
Art. 11(1) states that ENTSO-E (and ENTSOG – European Network
of Transmission System Operators for Gas) shall publish and submit to
the Member States, the Commission and the Agency their respective
methodologies, including on network and market modelling, for a harmonized
energy system-wide cost–benefit analysis at Union level for PCIs. Annex V
sets out the principles to be satisfied by the methodology for a harmonized
energy system-wide cost–benefit analysis for projects of common interest.
Art. 11(2) specifies that within three months of the day of receipt of the
methodologies, ACER shall provide an opinion to the Member States and the
Commission on the methodologies and publish it.
PCIs can then be ranked based on their CBA Annex III.2(11) of Regulation (EU) No 347/2013 setting out the process
and be selected via a process that is set out in the for establishing regional lists states that ‘The Group shall meet to examine
TEN-E Regulation. and rank the proposed projects taking into account the assessment of the
regulators, or the assessment of the Commission for oil and carbon dioxide
transport projects.’ Annex III.2(14) continues with ‘If, based on the regional
lists received, and after having taken into account the Agency opinion, the
total number of proposed projects of common interest on the Union list would
exceed a manageable number, the Commission shall consider, after having
consulted each Group concerned, not to include in the Union list projects
that were ranked lowest by the Group concerned according to the ranking
established pursuant to Article 4(4).’
Art. 4(4) of Regulation (EU) No 347/2013 states that ‘in order to facilitate
the assessing of all projects that could be eligible as PCIs and that could
be included in a regional list, each Group shall assess each project’s
contribution to the implementation of the same priority corridor or area
in a transparent and objective manner. Each Group shall determine its
assessment method on the basis of the aggregated contribution to the criteria
referred to in paragraph 2; this assessment shall lead to a ranking of projects
for internal use of the Group.’ Art. 4(4) continues with four additional criteria
to which each Group shall give due consideration when assessing projects.
82 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant article

regulation
Projects with PCI status have advantages, such Chapter III of Regulation (EU) No 347/2013 (Arts. 7–10) describes the
as accelerated permitting procedures, dedicated permit-granting and public participation procedure for PCIs. Chapter IV (Arts.
financial incentives from the NRAs involved, 11–13) describes their regulatory treatment. Lastly, Chapter V (Arts. 14–16)
and access to the Connecting Europe Facility specifies the financing. Art. 14 describes in more detail the eligibility criteria
(CEF). for Union financial assistance for projects. Recital 42 states that ‘Projects of
common interest in the fields of electricity, gas and carbon dioxide should be
eligible to receive Union financial assistance for studies and, under certain
conditions, for works as soon as such funding becomes available under the
relevant Regulation on a Connecting Europe Facility in the form of grants or
in the form of innovative financial instruments … A three-step logic applies
to investments in projects of common interest. First, the market should have
the priority to invest. Second, if investments are not made by the market,
regulatory solutions should be explored, if necessary the relevant regulatory
framework should be adjusted, and the correct application of the relevant
regulatory framework should be ensured. Third, where the first two steps
are not sufficient to deliver the necessary investments in projects of common
interest, Union financial assistance could be granted if the project of common
interest fulfils the applicable eligibility criteria.’
In the CBCA process, the TEN-E Regulation Art. 12(6) of Regulation (EU) No 347/2013 states that ‘Where the national
mandates ACER to act as a mediator in these regulatory authorities concerned have not reached an agreement on the
kinds of cases. TSOs make a CBCA proposal for investment request within six months of the date on which the request was
PCIs, and if the NRAs cannot agree, ACER can received by the last of the national regulatory authorities concerned, they
intervene. shall inform the Agency without delay. In this case or upon a joint request
from the national regulatory authorities concerned, the decision on the
investment request including cross-border cost allocation referred to in
paragraph 3 as well as the way the cost of the investments are reflected in
the tariffs shall be taken by the Agency within three months of the date of
referral to the Agency.’ Paragraph 3(c) states that the investment request shall
be accompanied by a substantiated proposal for cross-border cost allocation if
the project promoters agree.
PART II

How to combine electricity trade with system security to

keep the lights on?
5. Who is responsible for balancing the system?
Leonardo Meeus with Tim Schittekatte and Valerie Reif

In this chapter, we answer four questions. First, how to share responsibility between system
operators? Second, how to incentivize market parties to be balanced? Third, how to ensure that
reserves are available? And fourth, how to integrate balancing markets across borders?

5.1 HOW TO SHARE RESPONSIBILITY BETWEEN SYSTEM

OPERATORS?

In this section, we first explain why primary, secondary and tertiary control have been renamed
as the frequency containment, frequency restoration and reserve replacement processes. We
then discuss in more detail how the balance responsibility between system operators is shared
in each of these three processes. We also refer to a clock incident that illustrates the fragility
of the shared responsibility.

5.1.1 Name Change

The ‘primary’, ‘secondary’ and ‘tertiary’ control terminology is included in all power system
engineering handbooks. The various synchronous areas in Europe also applied this terminol-
ogy, but they had their own detailed definitions which did not align. The five synchronous
areas in Europe are Continental Europe (CE), the Nordics, the Baltics, Great Britain and
Ireland.
We worked as consultants for the European Commission in 2006 and were involved
in Babylonian discussions between experts from the Union for the Coordination of the
Transmission of Electricity (UCTE) and the Nordic Cooperation of Electricity Utilities
(NORDEL), the former bodies for cooperation between transmission system operators (TSOs)
in CE and the Nordics respectively. We did not recommend the name change, but we under-
stand why it became part of the solution. Table 5.1 introduces the new terminology and maps it
onto the historical concepts of primary, secondary and tertiary control. In the remainder of this
chapter, we use the new language of frequency containment, frequency restoration and reserve
replacement introduced in 2017 in the System Operation Guideline (SO GL).
Following the Electricity Balancing Guideline (EB GL), also introduced in 2017, a limited
number of standard balancing energy products were defined per balancing process. There is
one standard automatic frequency restoration reserves (aFRR) product, two standard manual
frequency restoration reserves (mFRR) products (direct and scheduled activation), and one
standard replacement reserves (RR) product.1 The Agency for the Cooperation of Energy
Regulators (ACER) was in favour of limiting the number of standard balancing energy prod-

84
 Who is responsible for balancing the system? 85

Table 5.1 Terminology for reserve products

Frequency containment Frequency restoration process Reserve

process replacement
process
Operational reserves Frequency containment Automatic frequency Manual frequency Replacement
defined in the SO GL reserves (FCR) restoration reserves (aFRR) restoration reserves reserves (RR)
(mFRR)
UCTE Primary control Secondary control Tertiary control
NORDEL Frequency-controlled
Fast reserves
reserves

Note: The mapping of names is indicative. Due to different operational practices a perfect one-on-one mapping
is not possible.

ucts during the EB GL drafting process, while the European Network of Transmission System
Operators for Electricity (ENTSO-E) proposal was to keep several products. The standardized
products, however, include standardized and non-standardized characteristics. A standardized
characteristic of a standard mFRR product is, for example, its full activation time and an
example of a non-standardized characteristic is the minimum duration between deactivation
and the following activation. Countries can also choose to introduce so-called specific balanc-
ing products, which are balancing services that can be used locally, to complement the stand-
ardized European products that will be exchanged across borders. To what extent countries
will use specific balancing products and to what extent the non-standardized characteristics of
the standard products will diverge is an open issue.

5.1.2 Frequency Containment and the Solidarity Mechanism

The frequency containment process is the first response to a frequency deviation when an
imbalance occurs. In each of the synchronous areas in Europe, this first response is an auto-
matic joint reaction by all the control areas in the synchronous area.
As is discussed in Box 5.1, frequency containment reserves (FCR) are dimensioned to
handle a so-called reference incident. Traditionally, this was the loss of the largest generation
unit. The probability that countries lost their largest unit at the same time was very low, so they
benefited from pooling their FCR requirements into a solidarity mechanism. Each synchro-
nous area already had a solidarity mechanism, and the SO GL formalized these mechanisms.
Without such a mechanism, each TSO responsible for a control area would have to procure
enough FCR to be able to handle the loss of its largest generator. For France, this would
imply doubling or even tripling the procured volume of FCR. For a smaller control area like
Belgium, it would imply increasing FCR procurement tenfold.
86 Evolution of electricity markets in Europe

BOX 5.1 PRIMARY FREQUENCY CONTROL IN UCTE AND

FREQUENCY-CONTROLLED RESERVES IN NORDEL

In UCTE, dimensioning primary control was based on a reference incident which defined
the maximum instantaneous power deviation between generation and demand to be han-
dled by primary control starting from undisturbed operation. This reference incident was
defined as 3000 MW, as was the primary control reserve in both the positive and negative
directions. This meant that a generation load imbalance of that size could be absorbed
without frequency deviations exceeding 200 mHz. The capacity chosen corresponded to the
simultaneous loss of two large units of around 1500 MW each. The provision of primary
control was a joint action involving generating units and loads spread evenly across the
interconnected network. The various countries’ contributions to the total primary regulation
reserve were determined on an annual basis and distributed in proportion to the share of the
energy generated in one year in the entire synchronous area.
In NORDEL, frequency-controlled reserves were split into normal operation reserves
and disturbance reserves. At least 600 MW of normal operation reserves needed to be
available at all times. These reserves were to be fully activated if the frequency deviation
from the nominal frequency exceeded 100 mHz. Contributions to the normal operation
reserves of the subsystems in the synchronous system were determined based on the annual
consumption in the previous year. The division was updated once a year. The volume and
composition of the disturbance reserves were dimensioned in such a way that a dimen-
sioning fault would not cause a frequency below 49.5 Hz in the synchronous system. The
disturbance reserves were distributed among the countries in proportion to their respective
dimensioning faults, for example the loss of a large generating station, and the division was
updated once a week, or more often if necessary.

Note: Box 5.1 is based on NORDEL (2006) and UCTE (2009). The contribution of each country to the FCR
solidarity mechanism can be found in Artelys (2017).

The definition of the reference incident is increasingly being questioned. Following the SO
GL, the reference incident for CE was defined as 3000 MW, which is the same as the historical
value shown in Box 5.1. The value is also the same for upward and downward FCR. In the
decentralizing power system the biggest generation unit has not increased, but we increasingly
have incidents involving new interactions between the smaller units, as will be shown in
Chapter 6. Whether this will lead to the definition of a new reference incident is an open issue.

5.1.3 Frequency Restoration and Reactive Balancing

The frequency restoration process starts directly after the frequency containment process. This
second response to a system imbalance is organized in so-called load frequency control (LFC)
areas. The LFC area is generally equal to the TSOs’ control area. Following the SO GL, TSOs
can decide to jointly operate the frequency restoration process in an LFC block spanning more
than one LFC area within a synchronous area. At the time of writing, most TSOs have not yet
done this. The restoration of an imbalance is therefore typically done by the relevant TSO in
the control area where the imbalance originates.
 Who is responsible for balancing the system? 87

Frequency restoration reserves (FRR) are used to restore the frequency within a predefined
time. First, automatic FRR (aFRR) and later manual FRR (mFRR) are activated to relieve FCR
and aFRR respectively. aFRR are automatically activated by a controller operated by the TSO;
mFRR are activated upon a specific manual request by the TSO.
Reactive balancing means that the TSO relies on market parties to balance themselves
through intraday markets. The TSO does not try to anticipate imbalances by activating
reserves before a frequency deviation takes place and the frequency containment and fre-
quency restoration processes kick in. This means that the balancing price can spike in periods
in which market parties have not managed to balance themselves in the intraday market. This
typically implies that there is much activity in the intraday market and market parties are
incentivized to help the system operator to balance the system. Most residual imbalances are
handled with aFRR.

5.1.4 Reserve Replacement and Proactive Balancing

The reserve replacement process involves the slowest type of reserves, which can need 30
minutes to be fully activated. The TSOs that have replacement reserves (RR) often rely less on
aFRR and apply a proactive balancing approach. They try to anticipate imbalances by activat-
ing RR before the frequency containment and restoration processes kick in.
Proponents of proactive balancing argue that it increases system security, which is espe-
cially important in more isolated and inflexible power systems. They also argue that the
cost of balancing is reduced by using cheaper RR instead of the more expensive FRR. The
counter-argument is that a proactive approach does not create enough incentives for market
parties to be balanced and invest in assets that can help balance the system. We will come back
to this later in this chapter, and also in Chapter 7 when we discuss the missing money problem.
The EB GL and the SO GL do not explicitly rule on proactive versus reactive balancing, but
they implicitly favour reactive balancing. A key concept in the codes is the balancing energy
gate closure time (GCT), which has been capped at one hour before real time in Europe.
Following the codes, market parties can submit balancing energy bids up to the GCT, and
TSOs cannot activate balancing energy bids prior to this GCT. This seems to limit TSOs to
balancing proactively, but there are exceptions. Several countries have also started to make use
of these exceptions to continue with proactive balancing. How this will evolve is very much
an open issue.

5.1.5 The Clock Incident

The fragility of the shared balance responsibility between system operators in Europe surfaced
in 2018. Many clocks depend on a stable frequency of 50 Hz to run correctly, but the average
frequency had been 49.996 Hz for several weeks. Clocks accumulated a delay of about six
minutes, which was the first time that electricity balancing reached the mainstream media.
ENTSO-E had to face the media to explain how a dispute in south-eastern Europe could
cause clock delays in the whole European continent.2 The dispute was between Serbia and
Kosovo in the LFC block of Serbia, North Macedonia and Montenegro (SMM block). The
system had been short, and had not been fully restored, with a total of 113 GWh missing.
88 Evolution of electricity markets in Europe

The dispute lasted several weeks. After it was resolved the TSOs decided to catch up by
maintaining a frequency of 50.01 Hz for some weeks. This corrected the delayed clocks,
except those that had already been restored manually, which were now running several
minutes ahead.

5.2 HOW TO INCENTIVIZE MARKET PARTIES TO BE

BALANCED?

In this section, we explain balance responsible parties (BRPs) and the imbalance settlement.

5.2.1 Balance Responsible Parties

In principle all market parties have balance responsibility. They can take this responsibility
themselves or delegate it to a third party. In what follows, we discuss exceptions and whether
this responsibility applies at the unit or portfolio level.
First, exceptions to balance responsibility. Renewable energy projects have often been
exempted from balance responsibility. They received feed-in tariffs so that they did not have
to worry about wholesale market prices or balancing prices. They just injected whatever they
could into the system and got paid for it. Forecasting was a problem for system operators, who
had to handle forecasting errors. There were no incentives for these new entrants to help in
that process. Following Regulation (EU) 2019/943, this is no longer possible. Only three types
of grid users can still receive a derogation: demonstration projects for innovative technologies
(limited in time); installations benefiting from support approved by the Commission under
Union state aid rules (commissioned before 4 July 2019); and renewable generation with an
installed electricity capacity of less than 400 kW (commissioned before 1 January 2026). For
renewable generation commissioned after 1 January 2026, the derogation can only apply to
installations with an installed electricity capacity of less than 200 kW.
Second, balance responsibility at the unit or portfolio level. The decentralized market model
in Europe entails that balance responsibility is defined at the portfolio level, typically within
a bidding zone. This gives market parties the possibility of deciding how to dispatch their
power plants. Some countries require market parties to balance their generation and consump-
tion portfolios separately, while other countries allow mixed portfolios. There are also excep-
tions. A few countries still have a more centralized market model, which is often combined
with central dispatch and balancing responsibility at the unit level. Note that with smaller
bidding zones (see the discussion in Chapter 3), the difference between portfolio bidding and
unit bidding becomes smaller. Even if there is portfolio bidding, with nodal pricing a BRP is
likely to only have a single unit or a very small portfolio in a certain node. If nodal pricing
were only applied to the transmission network, the portfolio would become the resources
connected to the distribution grid behind the transmission node.

5.2.2 Imbalance Settlement

In what follows we discuss the main design parameters in the imbalance settlement mecha-
nism: the imbalance settlement period and dual as opposed to single pricing.
 Who is responsible for balancing the system? 89

First, the imbalance settlement period (ISP). The ISP is the unit of time over which the
imbalance of a BRP is calculated. The ISPs in Europe range from 15 minutes (e.g. in the
Netherlands) to one hour (e.g. in Poland). This means that a BRP in Poland can be short in
some quarters of an hour and long in other quarters of the same hour without paying balancing
costs. In the Netherlands, the same BRP will face the costs for both imbalances. The more
BRPs face the costs they cause, the more they are incentivized to be balanced. The balancing
costs that cannot be allocated to BRPs are socialized by TSOs through their transmission
network tariffs. The EB GL states that all EU countries shall apply an ISP of 15 minutes within
three years of the entry into force of the regulation, although it leaves some room for excep-
tions. Regulation (EU) 2019/943 also states that the goal is a 15-minute ISP and adds that ISPs
shall not exceed 30 minutes from 1 January 2025, with no exceptions.
Second, dual versus single pricing. Dual pricing implies that BRPs that are unbalanced in
the direction of the system imbalance pay the costs of the balancing services that need to be
activated, often plus a penalty. The argument for the penalty is that market parties that are
unbalanced put the system in danger, so they need an extra incentive to avoid imbalances.
Dual pricing also implies that BRPs that are unbalanced in the opposite direction to the system
imbalance are not remunerated in the same way as the balancing services that are called upon
by the system operator. For instance, if a BRP is helping the system operator to solve a short-
age with its imbalance, its remuneration will be capped at the day-ahead price in wholesale
markets. Single pricing means no penalties and the same remuneration for activated balancing
services and BRPs that are unbalanced in the opposite direction to the system imbalance.
In our research we have long been advocating single pricing. Dual pricing is a legacy from
the past in which balancing was a mechanism rather than a market. Dual pricing increases
prices in wholesale markets because BRPs are incentivized to buy more than they need to
reduce the risk of being short in balancing markets and facing penalties. BRPs will also keep
reserves for themselves instead of offering them to the system operator. Dual pricing therefore
favours larger players that have reserves over new entrants that rely on the balancing markets.
We have also shown that dual pricing in one country can have negative consequences for
a neighbouring country because it incentivizes BRPs to move their imbalances across the
border to escape penalties.3
The EB GL clearly states that prices should reflect the real-time value of energy and favours
single pricing over dual pricing. However, following the EB GL, dual pricing is possible as
an exception. At the time of writing, some countries still use penalties, but their number has
decreased. Note that in terms of the geographical dimension, the imbalance price or prices
are calculated per imbalance price area. In this regard, Regulation (EU) 2019/943 states that
each imbalance price area shall be equal to a bidding zone. An exception to this is the case of
a central dispatching model, where an imbalance price area may constitute a part of a bidding
zone.

5.3 HOW TO ENSURE THAT RESERVES ARE AVAILABLE?

In this section, we first discuss balancing capacity tenders, then balancing energy markets, and
lastly the related issue of scarcity pricing in balancing markets.
90 Evolution of electricity markets in Europe

5.3.1 Balancing Capacity Tenders

In this subsection, we first discuss whether TSOs can own balancing services, and then discuss
how they tender for balancing capacity.
First, TSO ownership of balancing services. TSOs have in exceptional cases been allowed
to own pumped hydro plants and batteries. However, this creates a conflict of interest between
the TSO as the single buyer of balancing services and the TSO as one of the players in this
market. Directive (EU) 2019/944 ruled out TSO ownership of energy storage assets, with
possible derogations under specific circumstances. TSO ownership of generation assets has
already been addressed with the unbundling process, which we referred to in Chapter 1.
Second, TSO tenders for balancing capacity. Market parties that offer balancing services
are called balancing service providers (BSPs). They used to mainly be gas-fired power plants,
which preferred longer-term contracts and could provide upward and downward services in
one contract. This was nicely aligned with the interests of the system operators, who like
simplicity and security. Renewable energy players can more easily provide downward than
upward balancing services, and for aggregators of demand flexibility the opposite applies.
Both prefer separate procurements for upward and downward balancing services. They also
prefer tenders closer to real time, when they have a better idea of the services they can provide.
Some even prefer to offer balancing energy bids without balancing capacity reservation. In
some countries this was not possible; in other countries it was not appealing because whoever
won a balancing capacity tender had to provide balancing energy bids at the low price agreed
upon in the tender.4
To create a level playing field between incumbent and new players in balancing markets,
many changes were needed. They increased the complexity of balancing markets but also
improved their competitiveness and reduced the costs. This is an ongoing process. Regulation
(EU) 2019/943 contains provisions to gradually move away from yearly and monthly contracts
towards shorter-term day-ahead tenders. The EB GL states that the balancing energy price
should not be predetermined in the balancing capacity contract for standardized products.
This rule is reiterated in Regulation (EU) 2019/943. The EB GL also states that upward and
downward balancing capacity should be procured separately for FRR and RR. A temporary
exception to this rule can be requested. In Chapter 8, we will come back to the specific issues
related to flexible generation, demand and storage resources connected to distribution grids
and their participation in balancing markets.

5.3.2 Balancing Energy Markets

In what follows, we introduce the main balancing energy market controversies: merit-order
activation versus pro-rata activation; and marginal pricing versus pay-as-bid pricing.
First, merit-order activation versus pro-rata activation. As discussed above, many TSOs
relied on balancing capacity tenders to secure reserves. They did not always have a balancing
energy market to procure aFRR, mFRR or RR closer to real time. The price at which they acti-
vated reserves was often agreed upon in a balancing capacity contract. The contracted reserves
also contributed to solving an imbalance in proportion to their size without distinguishing
between delivery costs, that is, pro-rata activation. Under pro-rata activation, efficient pricing
and enabling competition for the provision of balancing energy is an issue. Therefore, the EB
 Who is responsible for balancing the system? 91

GL prescribes the development of balancing energy markets with merit-order activation of

balancing energy bids. Under merit-order activation, the cheapest balancing energy bids are
selected one by one up to the required volume. As described in the previous subsection, bal-
ancing capacity tenders can still be used to make sure that there are enough balancing energy
bids in the market, but not to set their price. Note that this applies to standardized balancing
products, but not necessarily to specific products that might still be used at the national level.
Second, marginal pricing versus pay-as-bid pricing. The EB GL makes marginal pricing
mandatory for standardized products. Pay-as-bid pricing is a legacy of pro-rata activation,
which has been replaced by merit-order activation. However, if all TSOs identify inefficien-
cies in the application of marginal pricing, they may request an amendment and propose an
alternative pricing method. Applying marginal pricing to balancing energy markets is not
always straightforward.5 Several bids can be activated at different times within a certain settle-
ment period, sometimes even in opposite directions. The EB GL did foresee the development
of a methodology for the harmonized implementation of marginal pricing in balancing energy
markets.
FCR is the exception. FCR continues to typically be a service that combines upward and
downward balancing. These reserves receive as much energy as they give, so that only capac-
ity needs to be remunerated. By its nature, the FCR product is indeed more about capacity than
energy.

5.3.3 Scarcity Pricing in Balancing Markets

Here, we first introduce the ‘missing money’ problem related to balancing capacity tenders,
and then discuss the scarcity pricing solution that has been proposed.6
First, the missing money problem. ACER and CEER’s market monitoring found that in
most of Europe balancing capacity is remunerated more than balancing energy, and imbalance
settlement prices only recover the balancing energy costs. This means that most balancing
costs are socialized through transmission network tariffs. It also means that BRPs are not ade-
quately incentivized to be balanced, and BSPs are not adequately incentivized to invest in flex-
ible assets. We already mentioned that this is an issue when discussing proactive balancing.
Balancing capacity tenders are an additional source of missing money in balancing markets.
Second, the solution. In Chapter 7 we will show that capacity mechanisms have been pro-
posed as a solution to compensate investors for the missing money problem. Another solution
is to reform balancing markets so that there is no missing money. The reform is referred to as
operating reserve demand curves (ORDC). It was proposed by academics and tested for the
first time in Texas in 2014. In Europe, the GB regulator Ofgem has applied a similar concept,
and the Belgian regulator CREG has started looking into it. The idea is to include an uplift
in the imbalance settlement pricing mechanism. When the system is short of reserves, the
uplift can increase to a very high value that represents the willingness of customers to avoid
being cut off from the system. When there are enough reserves in the system, the uplift goes
to zero. Note that the penalties we discussed in the previous section were also a kind of uplift.
However, they often disregarded system conditions and were designed to punish BRPs for
imbalances rather than to correct the scarcity price signal in balancing markets. The penalties
were also often combined with asymmetric incentives to BRPs and BSPs, while the uplift in
the ORDC approach applies symmetrically to BRPs and BSPs. The EB GL mentions that an
92 Evolution of electricity markets in Europe

additional settlement mechanism may be developed to allocate the costs of balancing capacity.
To what extent this will be done is an open issue.

5.4 HOW TO INTEGRATE BALANCING MARKETS ACROSS

BORDERS?

In this section, we first recall why the integration of balancing markets was started. We then
discuss the ongoing integration of balancing markets with imbalance netting, the exchange of
balancing energy, the exchange of balancing capacity and sharing of reserves. We end with
open issues in congestion management in balancing markets.

5.4.1 Why Was the Integration of Balancing Markets Started?

In the previous section, we discussed how ongoing changes to national balancing markets are
helping to bring new players into these markets. In this section, we discuss how the integration
of balancing markets across national borders is also helping to reduce balancing costs.
In Chapter 2 we referred to the Sector Inquiry published in 2007. We also discussed how
this inquiry helped to accelerate the integration of wholesale markets across borders. Table
5.2 shows why it did the same for balancing markets by revealing the alarming level of
concentration in national balancing markets between 2003 and 2005. Even countries that had
relatively well-functioning wholesale markets had problematic balancing markets. At the time
of the inquiry, TSO unbundling was also limited. The concern was therefore that TSOs were
not motivated enough to change the situation because they were in some cases part of the same
holding company as the generators profiting from the situation.
The integration process for balancing started after the wholesale process, but it proceeded
faster. The Nordics were the first to create a cross-border balancing market in Europe. Other
European TSOs followed with several pilots, and the ambition increased further during the
drafting process of the EB GL. The drafts referred to coordinated balancing areas (CoBAs).
They were going to be regional initiatives that would be slowly merged over time, like the
process that had been followed for wholesale markets as described in Chapter 2. ENTSO-E
was also proposing a minimalistic approach to balancing market harmonization. ACER argued
for harmonized products and integration through European platforms instead of CoBAs, and
the final version of the EB GL leans towards ACER’s view. Earlier in this chapter, we showed
that we have standardized products. In what follows, we introduce the European platforms.7

5.4.2 Imbalance Netting

As discussed earlier in this chapter, the first response to a frequency deviation is a shared
responsibility (i.e. frequency containment), but the second response (i.e. frequency restora-
tion) is generally the responsibility of the TSO from the control area that caused the frequency
deviation. This implies that one TSO might have a positive imbalance to solve while a neigh-
bouring TSO has a negative imbalance at the same time. Until recently, both TSOs would then
activate balancing services: one would balance upwards; the other downwards. Imbalance
netting simply means they stop doing this and only take action to solve the net imbalance.
The TSO with the residual net imbalance then still has to activate balancing services, but it
 Who is responsible for balancing the system? 93

Table 5.2 Market shares of the largest BSP for FRR between 2003 and 2005

UK ≈19% DK ≈88%
NL ≈39% FR ≈91%
ES ≈44% DE ≈93%
CZ ≈53% AT ≈100%
IT ≈78% HU ≈100%
SI ≈86%

Source: Author’s own table based on a graph from the Sector Inquiry (European Commission 2007).

is cheaper than solving the gross imbalance. The other TSO does not have to do anything.
Of course, netting can only take place if there is enough transmission capacity available to
accommodate the associated flows.
In our research, we found that the potential for Belgium and the Netherlands was significant.
The International Grid Control Cooperation (IGCC) project confirmed the benefits. The IGCC
is an initiative by the four German TSOs that grew into the biggest balancing pilot for imbal-
ance netting. More specifically, the IGCC avoids the activation of aFRR in opposite directions
on different sides of a border in the case that there is interconnector capacity available in the
right direction. Between 2012 and 2019, the accumulated benefit from this pilot was €475
million. And this with a cost of only €20 million for setting up the platform and an annual
cost of €1 million to run it. There were also smaller pilots running in parallel that confirmed
the potential of imbalance netting prior to the EB GL. However, it was later agreed that the
IGCC would be used as the starting point for the implementation of a European platform for
imbalance netting according to the EB GL. In January 2020, the IGCC was operational in 11
countries (13 TSOs). The initiative is quickly expanding as the EB GL mandates that the plat-
form shall be used to perform the imbalance netting process for at least the CE synchronous
area. The EB GL mandates that the platform shall be used to perform the imbalance netting
process for at least the CE synchronous area. It is foreseen that TSOs in synchronous areas
other than CE performing the automatic frequency restoration process (aFRP) may decide
to become member TSOs of the European platform at a later point in time. Netting strongly
reduces the need for balancing. As an illustration, in 2017, imbalance netting reduced Latvia’s
need for aFRR balancing energy by 55 per cent and the Netherlands’ need by 83 per cent.8

5.4.3 Exchange of Balancing Energy

The Nordics TSOs have been exchanging balancing energy for mFRR since 2002. More
recently, balancing pilots have also been set up by the other TSOs. The main aFRR pilot was
called EXPLORE (European X-border Project for LOng-term Real-time balancing Electricity
market design), and the RR pilot was called TERRE (Trans-European Restoration Reserves
Exchange).
Positive experience with these pilots inspired the EB GL drafters to set up European
platforms for the exchange of aFRR, mFRR and RR. The EB GL did foresee the develop-
ment of detailed implementation frameworks for each of these platforms, and TSOs have
already started with their establishment. The aFRR platform is called PICASSO (Platform
for the International Coordination of Automated Frequency Restoration and Stable System
94 Evolution of electricity markets in Europe

Operation). It was initiated in July 2017 and involved 16 TSOs (plus 10 observers) by the end
of 2019. PICASSO builds further on the work done in the EXPLORE regional project. The
mFRR platform is called MARI (Manually Activated Reserves Initiative). The RR platform is
called TERRE, which is the same name as the balancing pilot it is based on. TERRE is the only
balancing platform that has been launched as of January 2020. At that time, 8 TSOs (plus 6
TSO observers and ENTSO-E) were involved. In January 2020, ACER also adopted decisions
regarding the implementation frameworks for the other platforms.
Note that BSPs will continue to submit their balancing energy bids to the TSO of their
control area. The TSOs will then submit the bids to the European platforms. This way of
working is referred to as the TSO–TSO model. The EB GL states that this is the model for
exchanging balancing bids, with possible exceptions where two or more TSOs on their initia-
tive or on request by their relevant regulatory authorities develop a proposal for the application
of a TSO–BSP model. The alternative TSO–BSP model means that BSPs can submit their bids
to another TSO without passing through their own TSO. This then assumes that the BSPs can
purchase and use transmission rights for this purpose. The EB GL only allows alternative pro-
posals applying a TSO–BSP model as temporary solutions, except for certain circumstances
related to TSOs operating the replacement reserves process.

5.4.4 Exchange of Balancing Capacity and Sharing of Reserves

The EB GL requires each TSO to publish a balancing report at least every two years. In this
report the TSO is asked to analyse the potential for exchange of balancing capacity and sharing
of reserves and to put forward explanations and justifications for the procurement of balancing
capacity without the exchange of balancing capacity or sharing of reserves. This means that
much more information will be available, but the initiatives remain voluntary.
The exception is sharing of reserves for FCR, which is mandatory and already exists for
FCR, as was shown in Section 5.1.2. This implies that countries decide together the total
amount of FCR to procure and how to share the responsibility between them. In fulfilling this
responsibility, they can procure FCR locally or they can decide to exchange balancing capacity
across borders. A balancing pilot was set up between the Austrian TSO and the Swiss TSO to
exchange FCR capacity in 2013. This pilot was successful and currently the Belgian, Danish,
Dutch, French and German TSOs are also part of the FCR cooperation project. In 2019, the
project was formalized according to the EB GL.
Note that the SO GL required all the TSOs in a synchronous area to jointly develop a pro-
posal regarding the determination of LFC blocks. In many cases, the LFC block equals the
LFC area, which is equal to the control area. Exceptions are the Nordics, Ireland and Great
Britain, where the LFC block equals the synchronous area. Germany, Luxembourg and West
Denmark are also exceptions. They form an LFC block consisting of several LFC areas.
The same applies to Slovenia, Croatia, and Bosnia and Herzegovina and to Serbia, North
Macedonia and Montenegro. For these exceptions, the SO GL states that FRR has to be
dimensioned per LFC block. The TSOs involved then have to agree on the volume to reserve,
the ratio between aFRR and mFRR, and the sharing of responsibility. The SO GL also sets out
rules for dimensioning and sharing of RR for LFC blocks in which all the TSOs make use of
RR. These rules are similar to those for FCR, which apply to all TSOs. Following Regulation
(EU) 2019/943, regional coordination centres (RCCs) will facilitate regional dimensioning
 Who is responsible for balancing the system? 95

and the procurement of balancing capacity. RCCs were briefly introduced in Chapter 3 and
will be discussed in more depth in Chapter 6.

5.4.5 Congestion Management in Balancing Markets

There are three open issues in congestion management in balancing markets: reserving avail-
able transmission capacity for balancing markets; filtering balancing bids; and the interaction
with redispatching or flexibility markets.
First, the reservation of available transmission capacity for balancing markets. The EB GL
requires all TSOs to use available cross-zonal capacity after the cross-zonal intraday gate
closure for operating the imbalance netting process or for the exchange of balancing energy.
The EB GL also foresees the possibility of reserving available transmission capacity for the
exchange of balancing capacity or sharing of reserves for FRR and RR. Nothing is specified
regarding imbalance netting or the exchange of balancing energy. Note that available trans-
mission capacity is already reserved for FCR. The so-called reliability margin on transmission
lines is to account for FCR. To what extent it is opportune to do additional reservation is very
much an open issue.
Second, filtering balancing bids. The SO GL foresees distribution system operators (DSOs)
being able to filter balancing bids from BSPs connected to distribution grids. If they fear
that certain bids may cause congestion when called upon by TSOs, they can disqualify them.
Whether this will result in penalties being paid to the TSO by the BSP and/or paid to the BSP
or the TSO by the DSO is an open issue. The SO GL also allows TSOs to filter bids. They do
not have to forward all the BSPs’ bids to the European platforms. If they fear that certain bids
may cause congestion when called upon by neighbouring TSOs, they can also disqualify these
bids. Filtering balancing bids is a new issue that has emerged in the ongoing balancing market
integration process.
Third, the interaction between balancing markets and redispatching or flexibility markets.
Redispatching actions by TSOs were covered in Chapter 3. They imply that fewer resources
are available for balancing. Redispatching can be done by activating an upwards balancing
bid in one location and a downwards balancing bid in another location. It can also be done by
countertrading in the intraday markets. The TSO then procures energy in the intraday market
in one location to sell the same amount in another location. Some TSOs also procure redis-
patching resources in a separate market, possibly in cooperation with DSOs, which have also
started to use flexibility to manage congestion in their grids. How congestion management
in balancing markets, redispatching markets and flexibility markets will evolve is very much
an open issue. We will come back to the topic of TSO–DSO coordination in balancing and
congestion management in Chapter 8.9

5.5 CONCLUSION

In this fifth chapter on who is responsible for balancing the system, we have answered four
questions.
First, how to share responsibility between system operators? The first response to a fre-
quency deviation in Europe is referred to as the frequency containment process with frequency
containment reserves. This first response is also a joint response with a solidarity mechanism
96 Evolution of electricity markets in Europe

in each synchronous zone. The second response is referred to as the frequency restoration
process with automatic and manual frequency restoration reserves. This second response is
typically per control area and performed by the TSOs in the zone with the imbalance that
caused the frequency deviation. TSOs can also choose to share this responsibility by organiz-
ing themselves in so-called load frequency control areas and blocks that are bigger than one
control area. TSOs can anticipate frequency deviations and resolve imbalances before they
occur by activating RR, which is referred to as proactive balancing, as opposed to reactive
balancing. The EB GL default is reactive balancing, but proactive balancing is still possible.
Second, how to incentivize market parties to be balanced? All market players either take on
the balance responsible party role or delegate this responsibility to a third party. Renewable
energy players can no longer avoid this responsibility. The only exceptions are for very
small-scale and/or innovation demonstration projects. The imbalance settlement is becoming
more cost-reflective with single pricing instead of dual pricing, and with smaller imbalance
settlement periods.
Third, how to ensure that reserves are available? Many changes had to be made to balancing
markets to create a level playing field between the traditional balancing service providers and
new players. Balancing energy markets are becoming more important. At least for standard
energy balancing products, these markets will apply marginal pricing and the bids will be
activated following a merit order. Shorter-term balancing capacity tenders can continue. To
avoid the missing money problem, the costs of these tenders can be included in a scarcity price
signal through an uplift on top of the balancing energy price in periods when the system is
short of reserves.
Fourth, how to integrate balancing markets across borders? Imbalance netting has proven
to be very beneficial and relatively easy to implement. In line with the EB GL, TSOs have
started to set up European platforms to exchange balancing energy. Their names are PICASSO
(aFRR), MARI (mFRR) and TERRE (RR). Congestion management in balancing markets
is also an open issue. This includes reserving available transmission capacity for balancing,
filtering of balancing bids and the interaction between balancing markets and redispatching
and flexibility markets.

NOTES
1. There is no standard FCR balancing energy product. FCR is paid for reserved capacity rather than
activated volume.
2. More details can be found in the ENTSO-E press releases (2018a, 2018b, 2018c).
3. Saguan and Glachant (2007) were pioneers in the study of balancing market design, the negative
impact of dual pricing in balancing markets on wholesale prices and the willingness of balancing
service providers (BSPs) to pool reserves. In Vandezande et al. (2010) we elaborate on the negative
impact of dual pricing with penalties. We argue that the impact is not only discriminatory to new
entrant renewable energy players but also counterproductive in terms of the security of the system.
In Vandezande et al. (2009a) we illustrate how traders can exploit differences in the imbalance
settlement mechanisms of neighbouring countries. ENTSO-E (2019c) is an example of the annual
balancing market survey that is conducted by TSOs and includes the penalties that are applied in
countries that apply dual pricing.
4. De Vos et al. (2019) present a method for the dimensioning of operating reserves on a daily basis.
They show that this dynamic dimensioning approach makes a positive business case for the Belgian
system in 2020 as opposed to a static approach in which the capacity required is determined once
 Who is responsible for balancing the system? 97

a year. Based on the results of this study, Belgium decided on a gradual implementation towards
2020 of daily procurement of mFRR according to estimations of daily needs.
5. Littlechild (2007, 2015) provides a detailed discussion on the pricing rule for balancing energy.
He favours marginal pricing but also recognizes that some characteristics of the balancing energy
market make it more difficult to apply this rule in certain situations.
6. ACER first highlighted this issue in Figure 24 of its 2015 market monitoring report (ACER and
CEER 2016). Since 2017 (ACER and CEER 2018, 2019), the share of balancing costs recovered
through imbalance settlement has been left out of the report so it is difficult to know how it has
evolved. The academic who came up with the operating reserve demand curve (ORDC) concept
is William W. Hogan (Hogan 2005, 2013). Papavasiliou and Smeers (2017) study how an ORDC
adder would perform in Belgium. An evaluation of the implementation of scarcity pricing in GB can
be found in Ofgem (2018).
7. ENTSO-E (2019a) provides an overview of the nine balancing pilot projects in Europe. ENTSO-E
(2014) included the concept of CoBAs in version 3.0 of the EB GL, which was published in
2014. Pentalateral Energy Forum (2016) includes the difference between the views of ACER and
ENTSO-E on the level of harmonization required in balancing markets to enable integration during
the EB GL drafting process.
8. In Vandezande et al. (2009b) we calculate the potential for netting between Belgium and the
Netherlands using historical imbalance data, balancing prices and available transmission capacity.
The quarterly benefits of the IGCC are available in ENTSO-E (2019b), the costs related to the IGCC
in Artelys et al. (2016) and the data on Latvia and the Netherlands in ACER and CEER (2018).
9. The balancing survey in ENTSO-E (2019c) shows that in 2018 mFRR was also used for congestion
management in Great Britain and the Nordics. Balancing bids cannot be used for redispatch in the
Netherlands and Germany.

REFERENCES
50Hertz, Elia, Amprion, APG, Energienet, RTE, Swissgrid, Transnet BW and TenneT (2018), ‘TSOs’
Proposal for the Establishment of Common and Harmonised Rules and Processes for the Exchange
and Procurement of Balancing Capacity for FCR in Accordance with Article 33 of the EB GL’, pub-
lished on 18 October 2018.
ACER (2020a), ‘Decision No 01/2020 of the European Union Agency of 24 January 2020 on the
Methodology to Determine Prices for the Balancing Energy that Results from the Activation of
Balancing Energy Bids’.
ACER (2020b), ‘Decision No 02/2020 of the European Union Agency of 24 January 2020 on the
Implementation Framework for the European Platform for the Exchange of Balancing Energy from
Frequency Restoration Reserves with Automatic Activation’.
ACER (2020c), ‘Decision No 03/2020 of the European Union Agency of 24 January 2020 on the
Implementation Framework for a European Platform for the Exchange of Balancing Energy from
Frequency Restoration Reserves with Manual Activation’.
ACER and CEER (2016), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2015 – Electricity Wholesale Markets Volume’.
ACER and CEER (2018), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2017 – Electricity Wholesale Markets Volume’.
ACER and CEER (2019), ‘Annual Report on the Results of Monitoring the Internal Electricity and
Natural Gas Markets in 2018 – Electricity Wholesale Markets Volume’.
All CE TSOs (2018), ‘All TSOs’ Proposal for the Determination of LFC Blocks for the Synchronous
Area Continental Europe in Accordance with Article 141(2) of the SO GL’, published on 15 July
2018.
All Nordic TSOs (2018), ‘Proposal from all TSOs of the Nordic Synchronous Area for the Determination
of LFC Blocks within the Nordic Synchronous Area in Accordance with Article 141(2) of the SO GL’,
published on 5 July 2018.
98 Evolution of electricity markets in Europe

All TSOs (2018a), ‘All TSOs’ Proposal for the Implementation Framework for a European Platform for
the Exchange of Balancing Energy from Frequency Restoration Reserves with Automatic Activation
in Accordance with Article 21 of the EB GL’, published on 18 December 2018.
All TSOs (2018b), ‘All TSOs’ Proposal for the Implementation Framework for a European Platform for
the Exchange of Balancing Energy from Frequency Restoration Reserves with Manual Activation in
Accordance with Article 20 of the EB GL’, published on 18 December 2018.
All TSOs (2018c), ‘All TSOs’ Proposal on Methodologies for Pricing Balancing Energy and Cross-Zonal
Capacity Used for the Exchange of Balancing Energy or Operating the Imbalance Netting Process
Pursuant to Article 30(1) and Article 30(3) of the EB GL’, published on 18 December 2018.
All TSOs (2019), ‘All TSOs’ Proposal for the Implementation Framework for a European Platform
for the Imbalance Netting Process in Accordance with Article 22 of Commission Regulation (EU)
2017/2195 of 23 November 2017 Establishing a Guideline on Electricity Balancing’, published on 10
September 2019.
All TSOs using RR (2018), ‘The Proposal of all Transmission System Operators Performing the Reserve
Replacement Process for the Implementation Framework for the Exchange of Balancing Energy from
Replacement Reserves in Accordance with Article 19 of the EB GL’, published on 18 June 2018.
Artelys (2017), ‘METIS Technical Note T6 – METIS Power System Module’, METIS Technical Notes,
(May), accessed at https://​ec​.europa​.eu/​energy/​sites/​ener/​files/​power​_system​_module​.pdf.
Artelys, COWI and Frontier Economics (2016), ‘Integration of Electricity Balancing Markets and
Regional Procurement of Balancing Reserves’, Final Report, carried out for the European Commission
(DG ENER B1).
De Vos, K., N. Stevens, O. Devolder, A. Papavasiliou, B. Hebb and J. Matthys-Donnadieu (2019),
‘Dynamic dimensioning approach for operating reserves: Proof of concept in Belgium’, Energy
Policy, 124, 272–85.
ENTSO-E (2014), ‘ENTSO-E Network Code on Electricity Balancing Version 3.0’.
ENTSO-E (2018a), ‘Continuing frequency deviation in the Continental European Power System origi-
nating in Serbia/Kosovo: Political solution urgently needed in addition to technical’, press release of
6 March 2018.
ENTSO-E (2018b), ‘Deviations affecting frequency in Continental Europe have ceased; ENTSO-E
working on step 2’, press release of 8 March 2018.
ENTSO-E (2018c), ‘Frequency deviations – Continental European TSOs have restored the situation to
normal’, press release of 3 April 2018.
ENTSO-E (2019a), ‘Cross Border Electricity Balancing Pilot Projects’, accessed at https://​www​.entsoe​
.eu/​network​_codes/​eb/​.
ENTSO-E (2019b), ‘IGCC Regular Report on Social Welfare Q3 2019’.
ENTSO-E (2019c), ‘Survey on Ancillary Services Procurement and Balancing Market Design 2018’,
ENTSO-E WGAS, published in March 2019.
European Commission (2007), ‘DG COMP Report on Energy Sector Inquiry’.
Hogan, W. W. (2005), ‘On an “energy only” electricity market design for resource adequacy’, White
Paper, Harvard University.
Hogan, W. W. (2013), ‘Electricity scarcity pricing through operating reserves’, Economics of Energy and
Environmental Policy, 2 (2), 65–86.
Littlechild, S. (2007), ‘Electricity Cash Out Arrangements’, accessed at https://​www​.ofgem​.gov​.uk/​
ofgem​-publications/​40635/​19091cashoutreviewslittlechild​.pdf.
Littlechild, S. (2015), ‘Reflections on cash-out arrangements’, letter to Roger Witcomb, February 2015.
NORDEL (2006), ‘System Operation Agreement’.
Ofgem (2018), ‘Analysis of the first phase of the Electricity Balancing Significant Code Review’,
Working Paper, published in August 2018.
Papavasiliou, A. and Y. Smeers (2017), ‘Remuneration of flexibility using operating reserve demand
curves: A case study of Belgium’, Energy Journal, 38 (6), 105–36.
Pentalateral Energy Forum (2016), ‘Balancing’, Penta Flexibility Work Stream – Penta SG3 – Technical
Background Paper – Balancing.
Saguan, M. and J.-M. Glachant (2007), ‘An institutional frame to compare alternative market designs in
EU electricity balancing’, MIT-CEEPR Working Paper 07-001.
 Who is responsible for balancing the system? 99

UCTE (2009), ‘Continental Europe Operation Handbook. P1: Load-Frequency Control and Performance’.
Vandezande, L., L. Meeus, R. Belmans, M. Saguan and J. M. Glachant (2010), ‘Well-functioning balanc-
ing markets: A prerequisite for wind power integration’, Energy Policy, 38 (7), 3146–54.
Vandezande, L., L. Meeus, R. Belmans, M. Saguan, J. M. Glachant and V. Rious (2009a), ‘Lacking
balancing market harmonisation in Europe: Room for trader profits at the expense of economic effi-
ciency?’, 32nd IAEE International Conference, Jun 2009, San Francisco, United States.
Vandezande, L., M. Saguan, L. Meeus, J.-M. Glachant and R. Belmans (2009b), ‘Assessment of
the implementation of cross-border balancing trade between Belgium and the Netherlands’, 6th
International Conference on the European Energy Market, 27–29 May, Leuven, Belgium.
100 Evolution of electricity markets in Europe

5A.1 ANNEX: REGULATORY GUIDE

Table 5A.1 Regulatory guide

Section of this chapter, topic and relevant Relevant articles

regulation
Section 5.1.1
The SO GL introduces new names for old Art. 2(6–8) provides definitions of the terms ‘frequency containment
concepts: frequency containment, frequency reserves,’ ‘frequency restoration reserves’ and ‘replacement reserves’.
restoration, and reserve replacement.
There is one standard automatic frequency According to the EB GL, implementation frameworks have to be proposed
restoration reserves product (aFRR), two by all TSOs for the European aFRR, mFRR and RR balancing platforms
standard manual frequency restoration reserves (Arts. 21, 20, 19). One element in these implementation frameworks is
(mFRR) products (direct and scheduled proposed standard products (Art. 21(3.i), Art. 20(3.i), Art. 19(3.i)). Only the
activation) and one standard replacement implementation framework for RR has been approved by the relevant NRAs
reserves (RR) product. so far – in December 2018 (All TSOs using RR 2018). This framework
specifies that there is one standard RR product. The proposals for the
implementation frameworks for aFRR and mFRR were referred to ACER
in July 2019. In these proposals, All TSOs (2018a, 2018b) propose one
standard product for aFRR and two for mFRR. The mFRR standard products
differ in their activation mode, namely direct and scheduled activation. The
mFRR implementation framework proposals provide for balancing energy
bids to be submitted either only for scheduled activation or for scheduled and
direct activation in one common merit-order list. The mFRR implementation
framework defines a scheduled activatable bid in Art. 2(2.x) as ‘a standard
mFRR balancing energy product bid that can only be activated at one
specific point in time, i.e. the point of scheduled activation, with respect to
the period of time for which the balancing energy bid is submitted.’A direct
activatable bid is defined in Art. 2(2.e) of the same proposal as ‘a standard
mFRR balancing energy product bid that can be activated at any point of time
following the point of scheduled activation of the quarter hour for which the
bid is submitted and until the point of scheduled activation of the subsequent
quarter hour. Every direct activatable bid is a scheduled activatable bid as
well, while no scheduled activatable bid is a direct activatable bid.’
According to the EB GL, the standardized The standardized and non-standardized characteristics can be found in the
products include standardized and implementation frameworks for the respective balancing processes (Arts.
non-standardized characteristics. 21, 20, 19). The standardized characteristics of a product are defined in the
implementation frameworks. The non-standardized characteristics are defined
in the terms and conditions for balance service providers (BSPs), which are
defined by each TSO individually (Art. 18(1.a)).
 Who is responsible for balancing the system? 101

Section of this chapter, topic and relevant Relevant articles

regulation
According to the EB GL, countries can also Requirements for specific products are described in Art. 26. Art. 26(1.b and
choose to introduce so-called specific balancing f) states that a TSO proposing a specific product has to demonstrate that
products, which are balancing services that can standard products are not sufficient to ensure operational security and that
be used locally, to complement the standardized the specific product does not create significant inefficiencies and distortions
European products that will be exchanged across in the balancing market. Art. 26(3) states that ‘the specific products shall
borders. be implemented in parallel to the implementation of the standard products.
Following the use of the specific products, the connecting TSO may
alternatively: (a) convert the balancing energy bids from specific products
into balancing energy bids from standard products; (b) activate the balancing
energy bids from specific products locally without exchanging them.’ Art.
6(14) of Regulation (EU) 2019/943 reiterates these requirements for specific
products.
Section 5.1.2
Each synchronous area already had a solidarity Art. 153 describes the FCR dimensioning rules. Art. 153(1) states that ‘All
mechanism and the SO GL formalized these TSOs of each synchronous area shall determine, at least annually, the reserve
mechanisms. capacity for FCR required for the synchronous area …’
Following the SO GL, the reference incident Art. 153(2.b.i) states that ‘All TSOs of each synchronous area shall specify
for CE is defined as 3000 MW. The value is the dimensioning rules in the synchronous area operational agreement in
same for upward and downward FCR. accordance with the following criteria: … (b) the size of the reference
incident shall be determined in accordance with the following conditions:
(i) for the CE synchronous area, the reference incident shall be 3000 MW
in positive direction and 3000 MW in negative direction.’ Art. 153(2.c) adds
that ‘for the CE and Nordic synchronous areas, all TSOs of the synchronous
area shall have the right to define a probabilistic dimensioning approach for
FCR taking into account the pattern of load, generation and inertia, including
synthetic inertia as well as the available means to deploy minimum inertia in
real-time … with the aim of reducing the probability of insufficient FCR to
below or equal to once in 20 years.’
Section 5.1.3
According to the SO GL, the frequency Art. 141(4) states that ‘All TSOs of each LFC area shall … (b) implement and
restoration process (FRP) is organized in operate a FRP for the LFC area.’
so-called load frequency control (LFC) areas.
Following the SO GL, TSOs can decide to Art. 145(6) states that ‘In addition to the aFRP implementation in the LFC
jointly operate the frequency restoration process areas, all TSOs of an LFC block which consists of more than one LFC area
(FRP) in an LFC block spanning more than one shall have the right to appoint one TSO of the LFC block in the LFC block
control area in a synchronous area. operational agreement to: (a) calculate and monitor the FRCE of the whole
LFC block; and (b) take the FRCE of the whole LFC block into account for
the calculation of the setpoint value for aFRR activation in accordance with
Article 143(3) in addition to the FRCE of its LFC area.’In addition, an LFC
area can consist of more than one control area. If the LFC area is defined over
multiple control areas, TSOs can also jointly operate the FRP. An LFC area
is defined in the SO GL Art. 3(12) as ‘a part of a synchronous area or an
entire synchronous area, physically demarcated by points of measurement at
interconnectors to other LFC areas, operated by one or more TSOs fulfilling
the obligations of load-frequency control.’
102 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
Following the SO GL, frequency restoration Art. 143(1.a) states that ‘The control target of the FRP shall be to: (a)
reserves (FRR) are used to restore the frequency regulate the Frequency Restoration Control Error (FRCE) towards zero
within a predefined time. within the time to restore frequency.’ The FRCE is defined in Art. 3(43)
as ‘the control error for the FRP which is equal to the area control error
of a LFC area or equal to the frequency deviation where the LFC area
geographically corresponds to the synchronous area.’ Table 1 in Annex III
of the SO GL indicates that the time to restore frequency is 15 minutes for
each synchronous area, except for the Baltic synchronous area, for which no
information is listed.
Section 5.1.4
The replacement reserves process involves the The approved implementation framework for the exchange of balancing
slowest type of reserves, which can need 30 energy from replacement reserves states that the full activation time of the RR
minutes to be fully activated. standard product is 30 minutes. Art. 3(30) defines the full activation time as
‘the period between the activation request by the connecting TSO in case of
TSO–TSO model or by the contracting TSO in case of TSO–BSP model and
the corresponding full delivery of the concerned product.’
The EB GL and the SO GL do not explicitly There are several articles of importance with respect to this matter. Most
rule on proactive versus reactive balancing, but importantly, Art. 29(2–3) of the EB GL states that ‘2. TSOs shall not activate
they implicitly favour reactive balancing. The balancing energy bids before the corresponding balancing energy gate
balancing energy gate closure time is capped closure time, except in the alert state or the emergency state when such
at one hour before real time in Europe. Market activations help alleviate the severity of these system states and except
parties can submit balancing energy bids up to when the bids serve purposes other than balancing pursuant to paragraph
the GCT, and TSOs cannot activate balancing 3. 3. By one year after the entry into force of this Regulation, all TSOs shall
energy bids prior to this GCT. This seems to develop a proposal for a methodology for classifying the activation purposes
limit TSOs to balancing proactively, but there of balancing energy bids. This methodology shall: (a) describe all possible
are exceptions. purposes for the activation of balancing energy bids; (b) define classification
criteria for each possible activation purpose.’
Regarding the balancing energy gate closure time, two articles imply that
the balancing energy GCT for standard products will be at most one hour
before real time. First, with regard to the balancing energy GCT for standard
products, Art. 24(1–2) of the EB GL states: ‘1. As part of the proposals [for
the European platforms for the exchange of balancing energy], all TSOs shall
harmonize the balancing energy gate closure time for standard products at
the Union level, at least for each of the following processes: [RR, mFRR,
aFRR] 2. Balancing energy gate closure times shall: (a) be as close as
possible to real time; (b) not be before the intraday cross-zonal gate closure
time; (c) ensure sufficient time for the necessary balancing processes.’
Second, Art. 59(3) of the CACM GL concerning the intraday cross-zonal
GCT states that ‘One intraday cross-zonal gate closure time shall be
established for each market time unit for a given bidding zone border. It shall
be at most one hour before the start of the relevant market time unit and shall
take into account the relevant balancing processes in relation to operational
security.’ Note that the balancing gate closure time is only capped at one hour
for standard products of all reserve types. The timing of the gate closure for
specific products is not explicitly restricted.
 Who is responsible for balancing the system? 103

Section of this chapter, topic and relevant Relevant articles

regulation
Section 5.2.1
Following Regulation (EU) 2019/943, all market Art. 5(1) of Regulation (EU) 2019/943 states that ‘All market participants
participants are balance responsible with the shall be responsible for the imbalances they cause in the system (“balance
exception of three types of grid users which can responsibility”).’ Art. 5(2) states that ‘Member States may provide
still receive a derogation. derogations from balance responsibility only for: (a) demonstration projects
for innovative technologies, subject to approval by the regulatory authority,
provided that those derogations are limited to the time and extent necessary
for achieving the demonstration purposes; (b) power-generating facilities
using renewable energy sources with an installed electricity capacity of
less than 400 kW; (c) installations benefitting from support approved by the
Commission under Union State Aid rules pursuant to Articles 107, 108 and
109 TFEU, and commissioned before 4 July 2019.’ Art. 5(4) states that ‘For
power-generating facilities commissioned from 1 January 2026, point (b)
of paragraph 2 shall apply only to generating installations using renewable
energy sources with an installed electricity capacity of less than 200 kW.’
The decentralized market model in Europe Art. 54(1) states that ‘each TSO shall calculate within its scheduling area or
entails that balancing responsibility is defined scheduling areas when appropriate the final position, the allocated volume,
at the portfolio level, typically within a bidding the imbalance adjustment and the imbalance: (a) for each balance responsible
zone. Some countries require market parties party.’ Art. 110(2) states that ‘where a bidding zone covers only one control
to balance their generation and consumption area, the geographical scope of the scheduling area is equal to the bidding
portfolios separately, while other countries allow zone. Where a control area covers several bidding zones, the geographical
mixed portfolios. A few countries still have scope of the scheduling area is equal to the bidding zone. Where a bidding
a more centralized market model, which is often zone covers several control areas, TSOs within that bidding zone may jointly
combined with central dispatch and balancing decide to operate a common scheduling process, otherwise, each control
responsibility at the unit level. area within that bidding zone is considered a separate scheduling area.’
Furthermore, Art. 54(3) of the EB GL specifies that ‘Until the implementation
of the [proposal to further specify and harmonize the imbalance settlement],
each TSO shall calculate the final position of a balance responsible party
using one of the following approaches: (a) [the] balance responsible party
has one single final position equal to the sum of its external commercial
trade schedules and internal commercial trade schedules; (b) [the] balance
responsible party has two final positions: the first is equal to the sum of its
external commercial trade schedules and internal commercial trade schedules
from generation, and the second is equal to the sum of its external commercial
trade schedules and internal commercial trade schedules from consumption;
(c) in a central dispatching model, a balance responsible party can have
several final positions per imbalance area equal to generation schedules of
power generating facilities or consumption schedules of demand facilities.’
The possibility of keeping a central dispatch model is elaborated upon in the
following.
104 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
There are also exceptions. A few countries still The self-dispatch model is more in line with the European target model and
have a more centralized market model, which is seen as the ‘default’ in the EB GL, as can be deduced from Art. 14(2):
is often combined with central dispatch and ‘Each TSO shall apply a self-dispatching model for determining generation
balancing responsibility at the unit level. schedules and consumption schedules. TSOs that apply a central dispatching
model at the time of the entry into force of this Regulation shall notify to
the relevant regulatory authority … in order to continue to apply a central
dispatching model for determining generation schedules and consumption
schedules. The relevant regulatory authority shall verify whether the
tasks and responsibilities of the TSO are consistent with the definition [of
a central dispatching model] in Article 2(18).’ In EB GL Art. 2(18), a central
dispatching model is defined as ‘a scheduling and dispatching model where
the generation schedules and consumption schedules as well as dispatching of
power generating facilities and demand facilities, in reference to dispatchable
facilities, are determined by a TSO within the integrated scheduling process.’
Section 5.2.2
Balancing costs that cannot be allocated to Art. 44(2) of the EB GL states that ‘Each relevant regulatory authority in
BRPs are socialized by TSOs through their accordance with Article 37 of Directive 2009/72/EC shall ensure that all
transmission network tariffs. TSOs under its competence do not incur economic gains or losses with regard
to the financial outcome of the settlement pursuant to Chapters 2, 3 and 4 of
this Title [Settlement], over the regulatory period as defined by the relevant
regulatory authority, and shall ensure that any positive or negative financial
outcome as a result of the settlement pursuant to Chapters 2, 3 and 4 of this
Title shall be passed on to network users in accordance with the applicable
national rules.’
The EB GL states that all EU countries shall Art. 53 in the EB GL covers the imbalance settlement period. Art. 53(1) states
apply an ISP of 15 minutes by three years after that ‘By three years after the entry into force of this Regulation, all TSOs shall
the entry into force of the regulation, with some apply the imbalance settlement period of 15 minutes in all scheduling areas
room for exceptions. Regulation (EU) 2019/943 while ensuring that all boundaries of market time unit[s] shall coincide with
states that the goal is a 15-minute ISP and adds boundaries of the imbalance settlement period.’ Art. 53(2–3) explains the
that the ISP shall not exceed 30 minutes from 1 exemptions. Art. 8(4) of Regulation (EU) 2019/943 states that ‘By 1 January
January 2025, with no exceptions. 2021, the imbalance settlement period shall be 15 minutes in all scheduling
areas, unless regulatory authorities have granted a derogation or an
exemption. Derogations may be granted only until 31 December 2024. From
1 January 2025, the imbalance settlement period shall not exceed 30 minutes
where an exemption has been granted by all the regulatory authorities within
a synchronous area.’
 Who is responsible for balancing the system? 105

Section of this chapter, topic and relevant Relevant articles

regulation
The EB GL clearly states that prices should Art. 52(2) of the EB GL states that ‘By one year after entry into force
reflect the real-time value of energy and favours of this Regulation, all TSOs shall develop a proposal to further specify
single pricing over dual pricing. Following the and harmonize at least: … (c) the use of single imbalance pricing for all
EB GL, dual pricing is possible as an exception. imbalances … which defines a single price for positive imbalances and
negative imbalances for each imbalance price area within an imbalance
settlement period; and (d) the definition of conditions and methodology for
applying dual imbalance pricing for all imbalances … which defines one
price for positive imbalances and one price for negative imbalances for each
imbalance price area within an imbalance settlement period, encompassing:
(i) conditions on when a TSO may propose to its relevant regulatory authority
… the application of dual pricing and which justification must be provided;
(ii) the methodology for applying dual pricing.’ Additionally, Art. 6 of
Regulation (EU) 2019/943 states that ‘The imbalances shall be settled at
a price that reflects the real-time value of energy.’
Section 5.3.1
Directive (EU) 2019/944 rules out TSO Art. 54(1) states that ‘Transmission system operators shall not own,
ownership of energy storage assets, with possible develop, manage or operate energy storage facilities.’ Art. 54(2–5) lists
derogations under specific circumstances. the circumstances under which derogations from this rule can be granted by
Member States and the specific conditions that apply in such cases.
Regulation (EU) 2019/943 contains provisions to Art. 6(9) states that ‘… Contracts for balancing capacity shall not be
gradually move away from yearly and monthly concluded more than one day before the provision of the balancing capacity
contracts towards shorter-term day-ahead and the contracting period shall be no longer than one day, unless and to
tenders. the extent that the regulatory authority has approved the earlier contracting
or longer contracting periods to ensure the security of supply or to improve
economic efficiency.’ Art. 6(9–10) specifies the conditions that apply in cases
where a derogation is granted. Notwithstanding such derogations, Art. 6(11)
states that ‘from 1 January 2026 contract periods shall not be longer than six
months’ and Art. 6(12) states that ‘By 1 January 2028, regulatory authorities
shall report to the Commission and ACER on the share of the total capacity
covered by contracts with a duration or a procurement period of longer than
one day.’
The EB GL states that upward and downward Art. 32(3) states that ‘The procurement of upward and downward balancing
balancing capacity should be procured separately capacity for at least the frequency restoration reserves and the replacement
for FRR and RR. A temporary exception to this reserves shall be carried out separately. Each TSO may submit a proposal
rule can be requested. to the relevant regulatory authority … requesting the exemption to this
requirement.’ The article further lists what the proposal for exemption shall
include. Art. 6(9) of Regulation (EU) 2019/943 reiterates ‘The procurement
of upward balancing capacity and downward balancing capacity shall be
carried out separately, unless the regulatory authority approves a derogation
from this principle on the basis that this would result in higher economic
efficiency as demonstrated by an evaluation performed by the transmission
system operator …’
106 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
The EB GL states that the balancing energy price Art. 16(6) of the EB GL states that ‘The price of the balancing energy bids
should not be predetermined in the balancing or integrated scheduling process bids from standard and specific products …
capacity contract for standardized products. This shall not be predetermined in a contract for balancing capacity. A TSO may
rule is reiterated in Regulation (EU) 2019/943. propose an exemption to this rule … Such an exemption shall only apply to
A temporary exception to this rule can be specific products … and be accompanied with a justification demonstrating
requested. higher economic efficiency.’ Art. 6(2) of Regulation (EU) 2019/943 states that
‘The price of balancing energy shall not be pre-determined in contracts for
balancing capacity.’
Section 5.3.2
The EB GL prescribes the development of Arts. 19(2), 20(2) and 21(2) state that the European platforms for RR, mFRR
balancing energy markets with merit-order and aFRR respectively shall apply merit-order lists.
activation of balancing energy bids.
The EB GL makes marginal pricing mandatory Art. 30(1) states that ‘By one year after the entry into force of this Regulation,
for standardized products. all TSOs shall develop a proposal for a methodology to determine prices for
the balancing energy that results from the activation of balancing energy bids
for the frequency restoration process … and the reserve replacement process
… Such methodology shall: (a) be based on marginal pricing (pay-as-cleared
…’ Art.6(4) of Regulation (EU) 2019/943 states that ‘The settlement of
balancing energy for standard balancing products and specific balancing
products shall be based on marginal pricing (pay-as-cleared) unless all
regulatory authorities approve an alternative pricing method on the basis of
a joint proposal by all transmission system operators following an analysis
demonstrating that that alternative pricing method is more efficient.’
If all TSOs identify inefficiencies in the Art. 30(5) of the EB GL states that ‘Where all TSOs identify inefficiencies
application of marginal pricing, they may request in the application of the methodology proposed pursuant to paragraph 1(a),
an amendment and propose an alternative pricing they may request an amendment and propose a pricing method alternative to
method. the pricing method in paragraph 1(a). In such case, all TSOs shall perform
a detailed analysis demonstrating that the alternative pricing method is more
efficient.’
The EB GL did foresee the development All TSOs’ (2018c) proposal on methodologies for pricing balancing energy
of a methodology for the harmonized and cross-zonal capacity used for the exchange of balancing energy or
implementation of marginal pricing in balancing operating the imbalance netting process pursuant to Art. 30(1) and Art.
energy markets. 30(3) was referred to ACER in July 2019. In January 2020, ACER (2020a)
published its decision, which includes the implementation of a single
cross-border marginal price, which ensures that a single marginal price is
propagated across all areas among which there is no congestion. At least one
price is established for each market time unit, which is defined to be shorter
than or equal to the imbalance settlement period. The pricing methodology
also differentiates between the different products and processes.
FCR is an exception. It continues to typically be The EB GL states in Art. 32(3) that ‘The procurement of upward and
a service that combines upward and downward downward balancing capacity for at least the frequency restoration reserves
balancing. These reserves receive as much and the replacement reserves shall be carried out separately.’ Hence, this
energy as they give so that only capacity needs is not a requirement for FCR. Additionally, Art. 46(1) states that ‘each
to be remunerated. By its nature, the FCR connecting TSO may calculate and settle the activated volume of balancing
product is indeed more about capacity than energy for the frequency containment process with balancing service
energy. providers …’ Hence, paying BSPs for FCR is optional.
 Who is responsible for balancing the system? 107

Section of this chapter, topic and relevant Relevant articles

regulation
Section 5.3.3
The EB GL mentions that an additional Art. 44(3) states that ‘Each TSO may develop a proposal for an additional
settlement mechanism may be developed to settlement mechanism separate from the imbalance settlement, to settle the
allocate the costs of balancing capacity. procurement costs of balancing capacity … administrative costs and other
costs related to balancing.’
Section 5.4.2
The EB GL requires all TSOs that are Art. 22(1) states that ‘By six months after entry into force of this Regulation,
performing the automatic frequency restoration all TSOs shall develop a proposal for the implementation framework for
process to implement and make operational the a European platform for the imbalance netting process.’ At the time of
European platform for imbalance netting. The writing, the proposal by All TSOs (2019) has not yet been approved by
platform shall be used to perform the imbalance all NRAs. All NRAs made a second request for amendment in July 2019.
netting process at least for the Continental Furthermore, Art. 22(5) states that ‘By one year after the approval of the
Europe synchronous area. proposal for the implementation framework for a European platform for
the imbalance netting process, all TSOs performing the [aFR process] shall
implement and make operational the European platform for the imbalance
netting process. They shall use the European platform to perform the
imbalance netting process, at least for the Continental Europe synchronous
area.’
At the time of writing it is foreseen that TSOs Art. 5(4) of All TSOs’ revised proposal for the implementation framework
of synchronous areas other than CE performing for a European platform for the imbalance netting process from September
the automatic frequency restoration process may 2019 states that ‘TSOs of synchronous areas other than CE performing
decide to become member TSOs of the European the [automatic frequency restoration] process may decide to become
platform at a later point in time. member TSOs of the [imbalance netting] Platform at a later point in time,
after fulfilling the relevant requirements defined in this [implementation
framework] and the [imbalance netting] Platform accession roadmap.’
Section 5.4.3
The EB GL did foresee the development of According to the EB GL, implementation frameworks have to be proposed by
detailed methodologies for European platforms all TSOs for the European aFRR, mFRR and RR balancing platforms (Arts.
for the exchange of aFRR, mFRR and RR. 21, 20, 19). The implementation framework for RR was approved by the
relevant NRAs – in December 2018 (All TSOs using RR 2018). Proposals
for the implementation frameworks for aFRR and mFRR were referred
to ACER in July 2019. ACER (2020b, 2020c) adopted decisions on both
implementation frameworks in January 2020.
108 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
The EB GL states that the TSO–TSO model is Arts. 21, 20, 19 all state that the European balancing platforms shall apply
the model for exchanging balancing bids, with a multilateral TSO–TSO model with common merit-order lists to exchange
possible exceptions under certain circumstances. all balancing energy bids from all standard products. Art. 2(21) defines the
TSO–TSO model as ‘a model for the exchange of balancing services where
the balancing service provider provides balancing services to its connecting
TSO, which then provides these balancing services to the requesting TSO.’
Art. 33(2) states that ‘Except in cases where the TSO–BSP model is applied
pursuant to Article 35, the exchange of balancing capacity shall always be
performed based on a TSO–TSO model whereby two or more TSOs establish
a method for the common procurement of balancing capacity taking into
account the available cross-zonal capacity and the operational limits …’ Art.
35(1) states that ‘Two or more TSOs may at their initiative or at the request of
their relevant regulatory authorities … develop a proposal for the application
of the TSO–BSP model’ and lists what such a proposal for the application
needs to include.
The EB GL only allows alternative proposals as Art. 35(7) states that ‘By four years after entry into force of this Regulation,
temporary solutions. all exchanges of balancing capacity shall be based on the TSO–TSO model.
This requirement shall not apply to the TSO–BSP model for replacement
reserves if one of the two involved TSOs does not operate the reserve
replacement process as part of the load-frequency-control structure …’
Section 5.4.4
The EB GL requires each TSO to publish Art. 60(1) states that ‘At least once every two years, each TSO shall publish
a balancing report at least every two years. The a report on balancing covering the previous two calendar years, respecting
TSO is required to analyse the potential for the confidentiality of information in accordance with Article 11.’ Art. 60(2)
exchange of balancing capacity and sharing of specifies the information to be included in these reports, including ‘(e)
reserves and to put forward an explanation and analyse the opportunities for the exchange of balancing capacity and sharing
justification for the procurement of balancing of reserves; (f) provide an explanation and a justification for the procurement
capacity without exchanging balancing capacity of balancing capacity without the exchange of balancing capacity or sharing
or sharing reserves. of reserves.’
The FCR cooperation project was formalized In 2019, all relevant NRAs approved the methodology by all TSOs from the
according to the EB GL in 2019. Netherlands, Germany, France, Belgium, Austria, Denmark and Switzerland
on harmonized rules and processes for the exchange and procurement of
balancing capacity for FCR using one standard product according to Art. 33
of the EB GL (50Hertz et al. 2018).
The SO GL requires all TSOs in a synchronous Art. 141(2) of the EB GL states that ‘By 4 months after entry into force of this
area to jointly develop a proposal regarding the Regulation, all TSOs of a synchronous area shall jointly develop a common
determination of LFC blocks. In many cases, proposal regarding the determination of the LFC blocks, which shall comply
the LFC block equals the LFC area, which with the following requirements.’
equals the control area. The Nordics, Ireland The LFC block proposal by All CE TSOs (2018) was approved in August
and Great Britain, where the LFC block equals 2018. The LFC block proposal by All Nordic TSOs (2018) was approved in
the synchronous area, are exceptions. Germany, September 2018 by all the relevant NRAs.
Luxembourg and West Denmark, which form an
LFC block consisting of several LFC areas, are
also exceptions. The same applies to Slovenia,
Croatia and Bosnia and Herzegovina and to
Serbia, North Macedonia and Montenegro.
 Who is responsible for balancing the system? 109

Section of this chapter, topic and relevant Relevant articles

regulation
The SO GL states that FRR has to be Art. 157 states that all TSOs of an LFC block shall set out FRR dimensioning
dimensioned by LFC blocks. The TSOs involved rules in the LFC block operational agreement. The detailed requirements in
then have to agree on the volume to reserve, the terms of dimensioning are described in the article.
ratio between aFRR and mFRR, and the sharing
of responsibility.
The SO GL sets out rules for dimensioning and Art. 160 states that all TSOs of an LFC block shall have the right to
sharing of RR for LFC blocks in which all TSOs implement a reserve replacement process. The detailed requirements in terms
make use of RR. of dimensioning are set out in the article.
Regulation (EU) 2019/943 tasks the regional Art. 37(1.j and k) lists regional sizing of reserve capacity and facilitating
coordination centres (RCCs) with facilitating the regional procurement of balancing capacity as tasks of RCCs. Art. 6(7)
regional dimensioning and procurement of adds that ‘The dimensioning of reserve capacity shall be performed by the
balancing capacity. transmission system operators and shall be facilitated at regional level.’Art.
6(8) states that ‘The procurement of balancing capacity shall be performed by
the transmission system operator and may be facilitated at a regional level …’
Section 5.4.5
The EB GL requires that all TSOs shall use Art. 36(1) states that ‘All TSOs shall use the available cross-zonal capacity …
the available cross-zonal capacity after the for the exchange of balancing energy or for operating the imbalance netting
cross-zonal intraday gate closure for operating process.’Art. 37(3) specifies thatall the TSOs in a capacity calculation region
the imbalance netting process or for the shall develop a methodology for cross-zonal capacity calculation within the
exchange of balancing energy. balancing timeframe for the exchange of balancing energy or for operating the
imbalance netting process.
The EB GL foresees the possibility of reserving Art. 36(2.c) states that cross-zonal capacity for the exchange of balancing
available transmission capacity for the exchange capacity or sharing of reserves can be allocated. Art. 38(4) adds that
of balancing capacity or sharing of reserves for cross-zonal capacity allocated for the exchange of balancing capacity or
FRR and RR. Nothing is specified regarding sharing of reserves shall be used exclusively for FRR and RR. Art. 38(1)
imbalance netting or the exchange of balancing specifies three approaches allowed to reserve cross-zonal capacity for
energy. balancing: a co-optimization approach (Art. 40), a market-based approach
(Art. 41) and an approach based on economic efficiency analysis (Art. 42).
The capacity reserved shall be limited depending on the way it is calculated
(Art. 40(1.d), 41(2) and 42(2)). It should be noted that all TSOs exchanging
balancing capacity or sharing reserves shall regularly assess whether the
cross-zonal capacity allocated for the exchange of balancing capacity or
sharing of reserves is still needed for that purpose as it means that this
capacity is no longer offered to wholesale markets (Art. 38(8)).
Available transmission capacity is already Art. 38(4) states that ‘… The reliability margin calculated pursuant to
reserved for FCR. The so-called reliability [the CACM GL] shall be used for operating and exchanging frequency
margin on transmission lines is to account for containment reserves, except on Direct Current (DC) interconnectors
FCR. for which cross-zonal capacity for operating and exchanging frequency
containment reserves may also be allocated in accordance with paragraph
1.’ Paragraph 1 states that two or more TSOs may at their initiative or at the
request of their relevant regulatory authorities make a proposal for reserving
capacity for the exchange of balancing capacity according to one of the three
possible methods.
110 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
The SO GL foresees that DSOs can filter According to Art. 182(5), each reserve-connecting DSO and each intermediate
balancing bids from BSPs connected to DSO can set temporary limits to the delivery of active power reserves before
distribution grids. If they fear that certain bids their activation. Procedures need to be agreed upon with the respective TSO.
may cause congestion when called upon by Furthermore, the SO GL specifies in Art. 182(3) that the prequalification
TSOs, they can disqualify them. process for balancing resources connected to the distribution level shall rely
on rules concerning information exchanges and the delivery of active power
reserves between the TSO, the reserve-connecting DSO and the intermediate
DSOs. Each reserve-connecting DSO and each intermediate DSO, in
cooperation with the TSO, shall have the right to set limits to or exclude the
delivery of active power reserves located in the distribution system during
the prequalification process. Art. 182(4) specifies that reasons for limitations
or exclusion should be technical, such as the geographical location of the
reserve-providing units and reserve-providing groups.
Whether filtering bids for balancing by DSOs Art. 15(3) of the EB GL states that each TSO may, together with the
will result in penalties paid to the TSO by the reserve-connecting DSOs within the TSO’s control area, jointly elaborate
BSP and/or paid to the BSP or the TSO by the a methodology for allocating costs resulting from the exclusion or curtailment
DSO is an open issue. of active reserves connected to the distribution level.
The SO GL allows TSOs to filter bids. They Art. 29(14) on the activation of balancing energy bids from the common
do not have to forward all the bids by BSPs to merit-order list states that ‘Each TSO may declare the balancing energy
the European platforms. If they fear that certain bids submitted to the activation optimization function unavailable for
bids may cause congestion when called upon the activation by other TSOs because they are restricted due to internal
by neighbouring TSOs, they can also disqualify congestion or due to operational security constraints within the connecting
these bids. TSO scheduling area.’Additionally, Art. 29(10) states that ‘Each TSO
applying a self-dispatching model and operating within a scheduling area
with a local intraday gate closure time after the balancing energy gate
closure time pursuant to Article 24 may develop a proposal to limit the
amount of bids that is forwarded to the European platforms pursuant to
Articles 19 to 21. The bids forwarded to the European platforms shall always
be the cheapest bids. This proposal shall include: (a) the definition of the
minimum volume that shall be forwarded to the European platforms. The
minimum volume of bids submitted by the TSO shall be equal to or higher
than the sum of the reserve capacity requirements for its LFC block … and
the obligations arising from the exchange of balancing capacity or sharing
of reserves; (b) the rules to release the bids that are not submitted to the
European platforms and the definition of the point in time at which the
concerned balancing service providers shall be informed of the release of its
bids.’Art. 29(11) states that TSOs shall regularly assess the impact of limiting
the volume of bids sent to the European platforms and the functioning of the
intraday market.
6. How to organize system operation and
connection requirements?
Leonardo Meeus with Valerie Reif

In this chapter, we answer four questions. First, why pay attention to detailed technicalities?
Second, how to organize regional system operation? Third, how to introduce minimum techni-
cal standards? Fourth, how to proceed with the implementation of technical standards?

6.1 WHY PAY ATTENTION TO DETAILED TECHNICALITIES?

In this section, we revisit two iconic incidents where technical requirements came to the atten-
tion of the wider public. They involve a solar eclipse and a cruise ship.
First, the solar eclipse.1 On 20 March 2015, a solar eclipse took place that affected Europe,
starting at 08.01 in the western part of Portugal and ending at 11.58 in the eastern part of
Romania. It was the first solar eclipse in a power system with significant amounts of solar
power. Solar photovoltaic (PV) generation gradually reduces as the sun goes down in the
evening and raises back up with the sunrise. For the solar eclipse the same would happen, but
much faster. In anticipation of the event, regional groups of transmission system operators
(TSOs) in European Network of Transmission System Operators for Electricity (ENTSO-E)
had analysed the potential impact and had looked at countermeasures that could be useful
during the eclipse.
In the Nordics, the TSOs had not expected to be significantly affected by the reduction in
solar PV generation. The Nordic countries decided to limit exchanges with Continental Europe
(CE). The idea was to protect the Nordic system from a possible blackout in CE and also to
help CE recover from a blackout if this unlikely event were to occur.
Great Britain reported a combination of different effects during the solar eclipse. In compar-
ison with a normal day, there was a demand suppression due to people stopping their activities
to be able to watch the event. There was also a demand increase due to lighting that extended
beyond the eclipse. There was a loss of wind generation in addition to the loss of PV gener-
ation because solar eclipses typically reduce wind speeds. The balancing in real time by the
system operated mainly relied on pumped hydro storage.
In Continental Europe, the impact was expected to be the greatest. The installed solar PV
capacity in the CE synchronous area in 2015 was estimated at around 89 GW. As a protection
measure, some TSOs had increased their balancing reserves. The additional balancing costs
during the event ranged from €40 000 for France to €3.6 million for Germany. The Italian TSO
Terna had proactively removed 4400 MW of planned solar PV production from the day-ahead
market. TSOs had raised awareness and informed market players of the responsibilities they
had during the eclipse. Extra training of control room operators had been organized.
111
112 Evolution of electricity markets in Europe

During the eclipse, some TSOs strategically used pumped hydro storage plants.
Teleconferences among CE TSOs were held and continuous communication took place
between control rooms in CE and the Nordics. Ex post analysis showed that the frequency
quality during the event was high. In Continental Europe, the frequency never deviated more
than 48 mHz from the set point of 50 Hz. In Great Britain, the frequency only briefly exceeded
operational targets at the very beginning of the solar eclipse at 08.00.
Second, the cruise ship.2 The story starts with the Norwegian Pearl. This ship needed to
pass beneath a high-voltage line in the north of Germany during the evening of 4 November
2006. For security reasons, the line was manually disconnected by the TSO responsible. What
should have been a regular operational task that had been done several times during the pre-
vious years, on this day triggered the tripping of several high-voltage lines from the north to
the south of Continental Europe within a number of seconds. Lines that disconnected included
interconnection lines between the German TSOs E.ON Netz and RWE, internal E.ON Netz
lines (DE), internal APG lines (AT), interconnection lines between HEP (HR) and MAVIR
(HU), internal HEP lines (HR) and internal MAVIR lines (HU). Finally, the interconnection
lines between Morocco and Spain also tripped due to low frequency. The final report by UCTE
on the system disturbance concluded that the system was operated close to its security limits
and coordination and communication among the TSOs had been insufficient.
During this incident, the CE grid was split into three areas. Each experienced significant
power imbalances with a frequency drop in two of the areas and an increase in the third. The
TSOs completed full re-synchronization of the CE system 38 minutes after the splitting and
were able to re-establish a normal situation in all European countries in less than two hours.
Figure 6.1 illustrates the frequency deviations in the three CE areas, and the tripping of gener-
ation and shedding of load that followed.
The western area was composed of Spain, Portugal, France, Italy, Belgium, Luxemburg, the
Netherlands, a part of Germany, Switzerland, a part of Austria, Slovenia and a part of Croatia.
This area was facing a significant imbalance due to missing imports from the east, which led
to a quick drop in frequency down to about 49.0 Hz. At this frequency, larger generator units
synchronously connected to the transmission grid were expected to stay connected. In many

Source: Own illustration based on the UCTE (2007) report.

Figure 6.1 Capacities lost at the frequencies reached in the three areas during the split
of the Continental European system on 4 November 2006
 How to organize system operation and connection requirements? 113

countries, however, smaller generation units only connected to distribution grids had to be able
to withstand a drop to 49.5 Hz. In the western area, 10 900 MW of generation disconnected,
mainly wind and combined heat and power (CHP) generation connected to the distribution
grid. Except for one 700 MW thermal generation unit in Spain, none of the generation that
tripped was connected to the TSO networks. The load shedding that followed was aimed at
solving the initial imbalance in the western area plus the increased imbalance from the gener-
ation that disconnected. In total, about 1600 MW of pumps and 17 000 MW of consumption
were shed.
The north-eastern area was composed of a part of Austria, the Czech Republic, a part of
Germany, a part of Hungary, Poland, Slovakia and Ukraine. In this area, the frequency rose to
51.4 Hz due to a generation surplus of more than 10 000 MW. Automatic predefined actions
and automatic tripping of generation sensitive to high frequency values followed, mainly of
wind power connected to the distribution grid, which reduced the frequency to 50.3 Hz. The
tripping of wind generation was estimated at around 6200 MW and played a crucial role in
decreasing the frequency during the first seconds of the disturbance.
The south-eastern area was composed of North Macedonia, Montenegro, a part of Croatia,
Greece, Bosnia and Herzegovina, Serbia, Albania, a part of Hungary, Bulgaria and Romania.
This area only faced a slight under-frequency due to a lack of power of around 770 MW, but
no automatic actions or load shedding took place during the event.
The three main lessons learned can be summarized as follows. First, the cruise ship incident
was a wake-up call that there are many smaller units, like wind power plants, CHP and PV,
connected to the system that made the problem worse because they simply disconnected to
protect their own equipment. These units also caused challenges during frequency restoration,
as they automatically reconnected to the grid without any control by TSOs or DSOs when
the voltage and frequency conditions were in the acceptable range. In the new decentralized
system, it is not enough to manage the technical requirements of the larger units; the small
ones also have a technical responsibility that needs to be defined more clearly. This responsi-
bility cannot be limited to the national level because it will have consequences across borders.
Second, an understanding emerged that system operation needs to be regionalized. The cruise
ship incident was indeed the start of regional security coordination initiatives (RSCIs), which
were introduced to avoid this happening again. By sharing information and having a centre
that monitors different control areas, the national TSOs are better informed when taking
action. Third, the preparation for the first eclipse in a power system with a significant amount
of solar power worked out. Its ad hoc organization, however, triggered a debate on how such
coordination can be improved in the future.

6.2 HOW TO ORGANIZE REGIONAL SYSTEM OPERATION?

In Chapter 3, we showed how regional system operation evolved from RSCIs to regional
security coordinators (RSCs) to regional coordination centres (RCCs) following the cruise
ship incident discussed here. In this section, we first discuss in more detail the tasks of these
regional entities, which go beyond the capacity calculation task we focused on in Chapter 3.
We then introduce the Risk Preparedness Regulation (EU) 2019/941, which is part of the
Clean Energy Package (CEP) and was a response to the solar eclipse incident.
114 Evolution of electricity markets in Europe

6.2.1 Regional System Operation Tasks

Figure 6.2 shows how regional system operation tasks have evolved since voluntary initiatives
were launched in 2008. The voluntary RSCIs offered regional coordination services and
provided TSOs with an overview of electricity flows at the European regional level to comple-
ment the TSOs’ own system data. The System Operation Guideline (SO GL) later formalized
TSO coordination by setting out common requirements for the establishment of RSCs and for
their tasks.
RSCs deliver five core operational services to support national TSO decision-making,
namely coordinated capacity calculation, coordinated security analysis, delivery of the
common grid model, outage planning coordination, and short- and medium-term adequacy.
In addition, they are required to review the relevant TSO’s system defence and restoration
plans for consistency and to provide a technical report to all the TSOs involved, which in turn
transmit it to the relevant regulatory authorities and to ENTSO-E to monitor implementation.
RSCs also provide critical grid situation (CGS) support to TSOs, a responsibility that was
introduced with the cold spell in winter 2017/2018.3 This includes support to identify the risk
level at an early stage and to organize communication in advance and during the operational
planning phase if a CGS is expected and appropriate remedial actions needing to be prepared.
Furthermore, the SO GL defines certain annual reporting duties of RSCs to ENTSO-E.
The RCCs were created within the geographical scope of system operation regions (SORs),
a new concept introduced in the CEP. A SOR spans several established concepts, including
TSOs, bidding zones, bidding zone borders and capacity calculation regions. A SOR covers at
least one capacity calculation region and the TSOs in a SOR are required to participate in that
region’s RCC. All TSOs of a SOR are required to forward a proposal for the establishment of
RCCs to the relevant national regulatory authority (NRA) for approval by 5 July 2020. The
proposal should include the Member State in which the RCC will be located, the TSOs par-
ticipating, arrangements regarding the RCC’s organization, financing, operations and liability
together with an implementation plan for the RCC’s entry into operation. Following approval
by the NRAs, the RCCs will replace the RSCs and enter into operation by 1 July 2022.

Figure 6.2 Evolution of RSCIs into RSCs and then RCCs and a selection of their tasks
 How to organize system operation and connection requirements? 115

The five core operational services and the supporting tasks to ensure consistency of the
TSOs’ defence and restoration plans and manage CGSs will be transferred from the RSCs to
the RCCs. In addition, the portfolio of RCC tasks will include new tasks listed in the CEP.
In total, the CEP lists 16 tasks for RCCs. Some of the additional tasks are set by default,
for example post-operation and post-disturbance analysis and reporting, regional sizing of
reserve capacity and facilitation of regional procurement of balancing capacity. Other tasks
to be performed by RCCs are subject to a request from TSOs – for example, support for coor-
dination and optimization of regional restoration – or subject to delegation from ENTSO-E
– for example, identification of regional electricity crisis scenarios or carrying out seasonal
adequacy assessments. RCCs are to remain open to also extending their portfolio to potential
future needs for TSO regional coordination.
RCCs issue coordinated actions (for capacity calculation and security analysis) and recom-
mendations (for all other tasks) to the TSOs. The TSOs will implement coordinated actions
except where this would result in the violation of security limits. In such cases, the TSO is
required to file a report to the RCC and the other TSOs in the SOR. The RCC will assess
the impact of the decision on the other TSOs in the SOR and may propose a different set of
coordinated actions. If a TSO decides to deviate from a recommendation, it needs to submit
a justification to the RCC and the other TSOs in the SOR. Note that under certain conditions
the Member States in a SOR may jointly decide to grant their RCCs competence to issue coor-
dinated actions for one or more tasks for which currently only recommendations are given.

6.2.2 Risk Preparedness

The Risk Preparedness Regulation (EU) 2019/941, which was adopted as part of the Clean
Energy Package, sets out a common framework of rules on how to prevent, prepare for and
manage electricity crises. It considers that a regional approach brings significant benefits in
terms of the effectiveness of the measures implemented and the optimal use of resources, and
so requires Member States to cooperate in the regional context. Regulation (EU) 2019/941 was
inspired by the above solar eclipse incident.
As mentioned, RCCs are required to perform the tasks of regional relevance assigned
to them in accordance with both the Risk Preparedness Regulation (EU) 2019/941 and the
Electricity Regulation (EU) 2019/943. ENTSO-E can delegate to RCCs tasks relating to
seasonal adequacy assessment and to the identification of regional electricity crisis scenarios.
Note that delegated tasks are to be performed under the supervision of ENTSO-E. On the
basis of these regional electricity crisis scenarios, Member States will establish and update
their national electricity crisis scenarios, in principle every four years. These scenarios then
provide the basis for the national risk-preparedness plans that each Member State is required
to develop. In this regard, the geographical scope of SORs is relevant in the identification of
regional electricity crisis scenarios and risk assessments.

6.3 HOW TO INTRODUCE MINIMUM TECHNICAL

STANDARDS?

This section provides a high-level overview of the various requirements that are set out in
the connection network codes (CNCs). The CNCs have their basis in the network code areas
116 Evolution of electricity markets in Europe

referred to in Regulation (EC) No 714/2009. We first discuss the types of grid user and then
describe the types of requirement that apply to these grid users.

6.3.1 Different Types of Grid User

The two CNCs we cover in this chapter are the Requirements for Generators Network Code
(RfG NC) and the Demand Connection Network Code (DC NC). They refer to different
types of grid user: on the one hand power-generating modules (PGMs), which include syn-
chronous power-generating modules (SPGMs), power park modules (PPMs) and offshore
power park modules (OPPMs); on the other hand transmission-connected demand facilities,
transmission-connected distribution facilities, distribution systems, including closed distribu-
tion systems, and demand units which are used by a demand facility or a closed distribution
system to provide demand response services to relevant system operators. The different types
of grid user are explained in Annex 6A.1. Note that in this book we do not cover the third
connection code, the High Voltage Direct Current Network Code (HVDC NC).
The RfG NC applies to new PGMs which are considered significant, unless otherwise
provided. PGMs that are considered significant need to comply with the requirements of the
RfG NC according to the voltage level of their connection point and their maximum capacity
(in MW). As Table 6.1 shows, the RfG NC differentiates between four categories of PGMs
(types A to D). It only sets the upper limits (‘upper limit minimum capacity threshold’) of the
capacity thresholds used to divide the PGMs into different types, and these limits differ across
the synchronous areas. The final thresholds for the different types are set at the national level
and can be lower than the maximum threshold, except for type A PGMs. It is important to note
that the final thresholds chosen at the national level are also strongly dependent on national
practices before the entry into force of the RfG NC.
Generally, all the RfG NC requirements for lower category PGMs (e.g. PGM type A) need
to be fulfilled by those in a higher category (e.g. PGM type B). Note in this context that,
despite exceptions, the thresholds are generally lower in ‘smaller’ synchronous areas (with
a lower annual net electricity generation). This means that there are relatively more PGMs
in the higher classes (C and D), which generally have more stringent requirements to satisfy.
Furthermore, specific requirements apply to SPGMs of types B to D, PPMs of types B to D,
and AC-connected OPPMs. Last, the requirements can also differ between system operators
(DSOs or TSOs) when the specific requirement is to be set by the system operator that a grid
user has a grid connection contract with. All the requirements are minimum requirements,
which means that PGMs can always have more enhanced capabilities if this does not nega-
tively impact system security.
The DC NC splits the different requirements into two groups: first, requirements for
transmission-connected demand facilities, transmission-connected distribution facilities and
distribution systems, including closed distribution systems; second, requirements for demand
units used by a demand facility or a closed distribution system to provide demand response
services to relevant system operators. Depending on the voltage level of the connection, the
requirements can differ within these two groups.
 How to organize system operation and connection requirements? 117

Table 6.1 Limits for thresholds for different types of power-generating modules

Type Threshold
A ≥ 0.8 kW
Continental Europe Nordic Great Britain Ireland Baltic
B 1 MW 1 MW 1.5 MW 0.1 MW 0.5 MW
C 50 MW 50 MW 10 MW 5 MW 10 MW
D* 75 MW 75 MW 30 MW 10 MW 15 MW

Note: * PGMs with a connection point ≥ 110 kV are categorized as type D.

Source: RfG NC, Art. 5.

6.3.2 Different Types of Requirement

The technical requirements in the CNCs are divided into mandatory and non-mandatory
requirements, each of which can be either exhaustive or non-exhaustive. Exhaustive require-
ments do not need further national specification, while non-exhaustive requirements need
further specification at the national level for their entire application either in general or as
a site-specific choice, and within the boundaries set at the regional level. In other words,
non-exhaustive requirements do not provide for a full harmonization. Mandatory requirements
are to be applied in all EU Member States and other countries which implement the CNCs,
while for non-mandatory requirements countries can decide whether they want to introduce
such a requirement either in general at the national level or as a site-specific choice.4
Most of the requirements described in the RfG NC and DC NC are non-exhaustive
requirements. The specification of non-exhaustive requirements at the national level may
require updating and amending technical specifications, such as existing national grid codes.
Therefore, a transition period from the date of entry into force of the CNCs to their application
is foreseen, as is illustrated in Box 6.1.

BOX 6.1 RfG NC AND DC NC TIMELINES AND NATIONAL

IMPLEMENTATION PROCESSES

Figure 6.3 Timelines from publication to entry into application of RfG NC and DC NC
The Member States had the obligation to implement the CNCs at the national level no later
than three years after they were published, as is shown in Figure 6.3. Within this timeframe,
the relevant system operator, which in most cases was the TSO in coordination with the
118 Evolution of electricity markets in Europe

DSOs, had two years to define and submit a national specification of the non-exhaustive
requirements for approval by the competent entity. In order to support implementation at
the national level and in line with the legal requirements in the CNCs, ENTSO-E had an ob-
ligation to provide non-binding implementation guidance documents (IGDs). For reasons
of transparency, ENTSO-E monitored the CNC implementation process across the Member
States. Three years after the publication of the CNCs, all the parties affected had to com-
ply with the regulations. The CNC implementation process did not end with the entry into
application of the individual codes. On the contrary, all the relevant stakeholders have to
continually analyse the technical details of the codes and monitor whether the requirements
need to be revised in the light of system needs in future grid scenarios with increased pen-
etration of renewable energy sources.

Below we introduce the different types of non-exhaustive requirements related to frequency

issues, voltage issues, robustness, system restoration and general system management.5
First, there are three types of frequency-related requirements. The first type is related to
system inertia. As was explained in Chapter 1, the inertia of a system relates to the magnitude
of frequency deviations due to sudden imbalances between load and generation. If the system
inertia is low, a small sudden difference between load and generation causes a high frequency
deviation, and vice versa. Therefore, the RfG NC gives the relevant TSOs the right to specify
that PPMs of types C and D must be capable of providing synthetic inertia during very fast
frequency deviations to replace the effect of the inertia traditionally provided by SPGMs.
The level of inertia influences the frequency gradient (rate of change of frequency, RoCoF)
and therefore the stability of the power system. The RfG NC states that all types of PGM
must be capable of staying connected to the network and operating at RoCoFs up to a value
specified by the relevant TSO. In relation to this, the RfG NC and the SO GL address ramping
constraints. The DC NC states that demand units offering demand response must have the
withstand capability to not disconnect from the system under a RoCoF up to a value specified
by the relevant TSO.
A second type of frequency-related requirement specifies certain frequency ranges within
which PGMs, transmission-connected demand facilities, transmission-connected distribution
facilities, distribution systems and demand units providing demand response services should
be able to withstand frequency disturbances for a certain period of time without disconnecting
from the grid, as is shown in Table 6.2. This is important to avoid a sudden loss of a large
group of generators which could then initiate a cascading failure. The trade-off here is that
if PGMs or demand facilities are exposed to over-frequency or under-frequency for too long
their equipment can be damaged. Frequency ranges and duration of connection requirements
are defined by synchronous area. The RfG NC foresees that the relevant system operator
in coordination with the owner of the PGM may agree on wider frequency ranges, longer
minimum times for operation or specific requirements for combined frequency and voltage
deviations. The DC NC states that the owner of a transmission-connected demand facility or
the DSO may agree with the relevant TSO on wider frequency ranges or minimum times for
operation.
A third type of frequency-related requirement relates to the ability of PGMs and demand
units providing demand response services to contain or compensate for a frequency drop or
rise by regulating the active power output or input. This last type is related to the balancing
 How to organize system operation and connection requirements? 119

Table 6.2 Frequency ranges and duration of connection requirements by synchronous

area for PGMs of all types and transmission-connected demand facilities,
transmission-connected distribution facilities, distribution systems and
demand units providing demand response services

Frequency range Continental Nordic Great Britain Ireland Baltic

Europe
47.0–47.5 Hz - - 20 s - -
47.5–48.5 Hz Specified by each 30 min 90 min 90 min Specified by each
TSO, >=30 min TSO, >=30 min
48.5–49.0 Hz Specified by each Specified by each Specified by each Specified by each Specified by each
TSO, >=period for TSO, >=30 min TSO, >=90 min TSO, >=90 min TSO, >=period for
47.5–48.5 Hz 47.5–48.5 Hz
49.0–51.0 Hz Unlimited Unlimited Unlimited Unlimited Unlimited
51.0–51.5 Hz 30 min 30 min 90 min 90 min Specified by each
TSO, >=30 min
51.5–52.0 Hz - - 15 min - -

Source: RfG NC, Art. 13(1.a.i); DC NC, Art. 28(2.a), Art. 29(2.a) and Annex I.

mechanism (see Chapter 5) and includes requirements for both for operation under normal
conditions and for operation in emergency situations, such as operation in frequency sensitiv-
ity mode. Frequency sensitive mode is the PGM operating mode in which the active power
output changes in response to a change in the system frequency so as to assist with recovery
to the target frequency.
Second, there are three types of voltage-related requirements. An important difference
with respect to frequency is that voltage issues have to be dealt with locally, although they
can also have a cross-border impact if not dealt with properly. Frequency can be controlled
by adjusting active power consumption or generation, while voltage is controlled by reactive
power consumption or generation. A first type of voltage-related requirements is reac-
tive power requirements. The DC NC specifies these for transmission-connected demand
facilities and transmission-connected distribution systems. A second type specifies certain
voltage ranges within which PGMs of type D, transmission-connected demand facilities,
transmission-connected distribution facilities and transmission-connected distribution systems
should remain connected to the grid for certain time periods. As with frequency ranges and
their duration, the time period chosen for remaining connected during a voltage deviation
should be long enough for the TSO to take the necessary mitigating actions and short enough
to limit constraints on grid user equipment. A third type of voltage-related requirements is
reactive power capabilities of PGMs of types B to D and demand units providing demand
response services to contain or compensate for voltage deviations from the reference values.
As with frequency, this last type covers both actions taken under normal system operation and
actions taken in emergency situations.
Third, there are robustness-related requirements. Robustness or resilience can be defined as
the ability to cope with disturbances without loss of proper functioning. In the context of the
CNCs, this means the ability to cope with disturbances and to help prevent any major disrup-
tion or to facilitate restoration of the system after a collapse. More specifically, generation and
120 Evolution of electricity markets in Europe

Table 6.3 Compliance of different types of power-generating modules with a selection

of technical requirements as described in the RfG NC

Type A Type B Type C Type D

Frequency ranges x x x x
Fault-ride-through capability x x x
Operation in frequency sensitivity mode x x
Voltage ranges x

demand units are required to remain connected to the grid after a sudden voltage dip to help
prevent any major disruption or to facilitate restoration of the system after a collapse. Voltage
dips in high-voltage transmission grids are caused by switching activities which result in
a redistribution of energy flows happening as a result of a short circuit or a planned disconnec-
tion. The RfG NC sets out requirements for PGMs of types B to D to remain connected to the
network and operate through periods of low voltage at the connection point caused by secured
faults. This is the so-called fault-ride-through capability. Moreover, SPGMs of types B to D
are required to contribute to minimizing the spread of a voltage dip by recovering their active
power output quickly following a voltage disturbance. Similar specifications exist for PPMs
of types B to D. Relevant in this regard are also the requirements set out in the DC NC for
transmission-connected demand facilities and transmission-connected distribution systems to
withstand high short-circuit currents to avoid a potential cascading disconnection of grid users.
Fourth, there are system-restoration-related requirements. In Section 6.2, we introduced the
system defence and restoration plans that TSOs are required to establish. It is important to have
PGMs that do not need any external supply of electrical energy to restart after a blackout takes
place. This is referred to as black start capability. Most PGMs do not have this capability. The
RfG NC does not mandate black start capability for PGMs but refers to the right of Member
States and TSOs to require owners of type C and D PGMs to equip their PGMs with a black
start capability under certain conditions. Next to black start capability, the RfG NC also
includes system-restoration-related requirements as regards island operation, reconnection and
re-synchronization. For all three types of requirements related to system restoration, the CNCs
set out some obligations while other requirements are to be defined by the relevant system
operator.
Fifth, there are general system management related requirements. These include require-
ments related to information exchange with the relevant system operator, the settings of
control and protection schemes, instrumentation needed to provide fault recording and monitor
dynamic system behaviour and simulation models needed to test PGM compliance with the
different technical requirements. These requirements can be different for different types of
PGMs and for generation and demand units.
Note that the higher the classification of the PGM, the more requirements it has to satisfy,
and the more stringent these requirements are, which is illustrated in Table 6.3. We will come
back to this when we discuss open issues.
 How to organize system operation and connection requirements? 121

6.4 HOW TO PROCEED WITH THE IMPLEMENTATION OF

TECHNICAL STANDARDS?

In this section, we first discuss ongoing CNC implementation issues and then look at the case
of energy storage.

6.4.1 Implementation Issues

Here we discuss four implementation issues regarding CNCs: the compliance by existing
versus new connections, multiple connections versus a single connection, derogations at the
national level and requirements versus markets.
A first open issue is compliance by new versus existing connections. The DC NC applies to
new transmission-connected demand facilities, new transmission-connected distribution facil-
ities, new distribution systems, including closed distribution systems, and new demand units
used by a demand facility or a closed distribution system to provide demand response services
to relevant system operators. As mentioned in Section 6.3.1, the RfG NC applies to PGMs that
are new and significant. Significant means a PGM capacity larger than or equal to 0.8 kW.
PGMs and transmission-connected demand facilities, transmission-connected distribution
facilities, distribution systems and demand units used to provide demand response services
are considered new if the final and binding contract for the purchase of the main generating
plant, demand equipment or demand unit is not finalized by two years after the entry into force
of the relevant regulation. However, the RfG and the DC NC can also exceptionally apply to
existing connections. A first exception is that TSOs can propose to the relevant regulatory
authority to make existing PGMs, existing transmission-connected demand facilities, existing
transmission-connected distribution facilities, existing distribution systems or existing demand
units subject to all or some of the requirements of the relevant regulation. Such retroactive
action can be done if the TSO deems that it faces significant factual changes in circumstances,
such as an evolution of system requirements (including the penetration of renewable energy
sources, smart grids, distributed generation or demand response), and the regulator agrees with
it. A second exception is that, after notification by the system operator, the relevant regulatory
authority can decide that existing PGMs of types C or D, existing transmission-connected
demand facilities, existing transmission-connected distribution facilities, existing distribution
systems or existing demand units have been modernized substantially and therefore need
a revised or new connection agreement. How this will be handled at the Member State level
remains to be seen.
A second open issue is multiple connections versus a single connection. As Table 6.3
shows, the classification of different types of PGMs impacts on how stringent the require-
ments are that the PGM has to comply with. A key question is whether a project can request
multiple connections. Consider a wind park with ten 3.5 MW wind turbines. If the project is
treated as a single 35 MW connection, it will be classified as a type C or even type D gen-
eration unit, depending on the control area. If it is treated as ten 3.5 MW connections, they
will be classified as ten type B generation units. In the second scenario, the project will have
to comply with less stringent requirements. The RfG NC states that classification should be
based on size and the effect on the overall system. Differences exist between synchronously
and non-synchronously connected generation units. For synchronous generators, the RfG NC
122 Evolution of electricity markets in Europe

states that the PGM capacity should be based on machine size and include all the components
of a generating facility that normally run indivisibly. This means that one large turbine equals
one generation unit. For non-synchronous generation units, the RfG NC is less clear. It states
that non-synchronously connected PGMs should be assessed on their aggregated capacity
where they are collected together to form an economic unit and where they have a single con-
nection point. It all comes down to the definition of economic unit and single connection point,
which is left to Member States, so how this will be handled in detail is an open issue. Note
that the DC NC does not define different categories with different technical requirements, so
this issue does not apply to demand connections. However, the DC NC clarifies that demand
connections with more than one demand unit are to be considered as one demand unit if they
cannot be operated independently from each other or can reasonably be considered in a com-
bined manner.
A third open issue is derogations at the national level. Three aspects of derogations are
worth noting. First, the criteria for granting derogations are defined by the NRAs. The
European Commission may require an NRA to amend the criteria if they are not in line with
the CNC. Additionally, if an NRA deems that it is necessary due to a change in circumstances
relating to an evolution of system requirements, it may review and amend the criteria for
granting derogations at most once a year. Second, two types of derogations exist depending
on who files the request. Derogations can be requested by a PGM or demand facility owner
or prospective owner. They can also be requested by a system operator for classes of PGMs
or for multiple demand facilities. In such a case, the system operator would consider the need
for more stringent requirements than those provided by the CNC to guarantee secure system
operation. Third, all decisions regarding derogations granted or refused are notified to ACER
and kept in a regularly updated register. ACER and the European Commission both have the
possibility of issuing a reasoned recommendation to a regulatory authority to revoke a deroga-
tion due to a lack of justification. It remains to be seen to what extent derogations will be used.
A fourth open issue is requirements versus markets. Non-frequency ancillary services may
follow a similar path to frequency ancillary services. As was discussed in Chapter 5, frequency
ancillary services, or balancing, evolved from technical mechanisms into markets. As discussed
in this chapter, non-frequency ancillary services can be treated as technical requirements in
connection agreements, but they can also evolve into services that are procured in markets.
In some countries, TSOs have already started to define and procure services such as black
start capabilities and voltage control. The power system of the future might also require the
definition of new requirements and/or services. Inertia might also become a scarce resource in
larger synchronous systems as we move towards a power system with more non-synchronous
generation units. Regulation (EU) 2019/943 foresees a second generation of network codes
which will include rules on the non-discriminatory transparent provision of non-frequency
ancillary services, including rules on voltage control, inertia, black start capability and others.
To what extent these rules will lead to new requirements or markets is an open issue.6

6.4.2 The Case of Energy Storage

Two aspects of current discussions around energy storage are worth noting. First, with the
exception of pump-storage power-generating modules, which were included in the RfG NC,
the first generation of network codes does not cover energy storage. Member States are
 How to organize system operation and connection requirements? 123

therefore free to follow one of three paths. They can treat storage under one of the existing
asset classes of generation or demand and apply the relevant requirements of the relevant
connection code; they can create a new asset class and create a separate grid code for energy
storage; or they can also do something in between by starting from one of the existing codes
and making a few adaptations to reflect the specificities of storage technologies.
Second, Regulation (EU) 2019/943 foresees a second generation of network codes to
specify rules in relation to demand response, including rules on aggregation, energy storage
and demand curtailment. We do not yet know whether these rules will be more like market
guidelines alongside the existing CACM GL, FCA GL and EB GL or whether they will
become grid connection network codes alongside the RfG NC, DC NC and HVDC NC.

6.5 CONCLUSION

In this chapter on how to organize system operation and connection requirements we have
answered four questions. First, why pay attention to detailed technicalities? The technical
requirements of grid connections came to the attention of the general public when we had
a near-blackout experience triggered by a cruise ship in Germany. The first lesson learned
was that system operation needs to be regionalized. This was indeed the start of voluntary
initiatives that were later formalized in the SO GL and raised to the next level with the Clean
Energy Package. The second lesson learned was that we need stricter requirements for new
smaller units connected to the power system. The preparations for the first solar eclipse in
a power system with significant amounts of solar power also inspired a more formal risk
impact assessment process at the European level.
Second, how to organize regional system operation? TSOs remain the only system oper-
ators at the transmission level, but the roles of regional entities have been increasing with
the changes from RSCIs to RSCs and RCCs. The Risk Preparedness Regulation, which was
adopted as part of the Clean Energy Package, introduced a regional approach for preventing,
preparing for and managing electricity crises, such as solar eclipses.
Third, how to introduce minimum technical standards? The connection network codes set
out technical requirements for connecting different users and technologies to the grid. The
Member States were obliged to implement the connection network codes at the national level
no later than three years after their publication. Many non-exhaustive technical requirements
had to be defined and tough choices had to be made. The requirements will continue to evolve
with the power system.
Fourth, how to proceed with the implementation of technical standards? The new technical
requirements mainly apply to new connections, unless the TSO successfully argues that the
system needs a retroactive intervention. TSOs and grid users can also apply for derogations
from the requirements that have been set, and grid users might try to escape some of the
requirements by applying for multiple connections instead of a single bigger connection for
a certain project. Requirements could also be at least partly replaced by markets remunerating
grid users for the non-frequency ancillary services they provide. These are all open issues. The
second generation of network codes might also play a role, especially in energy storage, which
has not yet been addressed by the first generation of codes.
124 Evolution of electricity markets in Europe

NOTES
1. This account of the solar eclipse is mostly based on the impact analysis by ENTSO-E (2015) and
a report by ENTSO-E and SolarPower Europe (2015).
2. The incident is discussed in detail in UCTE (2007). See also the report in German by the German
national regulatory authority Bundesnetzagentur (2007). Researchers from the Ecole Centrale de
Lille (FR) published a video of the event, which is available on YouTube under the title ‘System
Disturbance in EUROPE (2006)’.
3. The RSC task of critical grid situation management is not mandated by the network codes but was
introduced by ENTSO-E (2018, 2019) to better tackle cold spells during winter.
4. ENTSO-E (2016a) explains the differentiation and the relationship between mandatory/
non-mandatory and exhaustive/non-exhaustive requirements in a guidance document for national
implementation of the CNCs.
5. Note that we do not differentiate between mandatory and non-mandatory non-exhaustive require-
ments. In its guidance document, ENTSO-E (2016b) gives an overview of non-exhaustive require-
ments in the CNCs and specifies which parameters are mandatory and which are non-mandatory.
6. For a long time there has been a debate in academia on whether in the presence of voltage con-
straints reactive power prices should also be determined in addition to active power prices. For
example, Hogan (1993) claims that reactive power prices complementing active power prices are
needed, while Kahn and Baldick (1994) state that ‘reactive power is cheap’, and that reactive power
pricing can be very hard in practice. Anaya and Pollitt (2018) explore the international experience of
system operators procuring reactive power in different jurisdictions, including Australia, the United
States and Great Britain.

REFERENCES
Anaya, K. L. and M. G. Pollitt (2018), ‘Reactive power procurement: Lessons from three leading coun-
tries’, Cambridge Working Papers in Economics: 1854.
Bundesnetzagentur (2007), ‘Bericht der Bundesnetzagentur für Elektrizität, Gas, Telekommunikation,
Post und Eisenbahnen über die Systemstörung im deutschen und europäischen Verbundsystem
am 4. November 2006’ [‘Court of Justice of the European Communities for Electricity, Gas,
Telecommunications, Post and Electricity Transmission Systems in the European Union and in the
European Union on 4 November 2006’].
ENTSO-E (2015), ‘Solar Eclipse 2015. Impact Analysis’.
ENTSO-E (2016a), ‘Making Non-Mandatory Requirements at European Level Mandatory at National
Level. ENTSO-E Guidance Document for National Implementation for Network Codes on Grid
Connection’.
ENTSO-E (2016b), ‘Parameters of Non-Exhaustive Requirements. ENTSO-E Guidance Document for
National Implementation for Network Codes on Grid Connection’.
ENTSO-E (2018), ‘Annual Report 2017’.
ENTSO-E (2019), ‘Enhanced TSO Regional Coordination for Europe: Act Locally, Coordinate
Regionally, Think European’.
ENTSO-E and SolarPower Europe (2015), ‘Solar eclipse March 2015: The successful stress test of
Europe’s power grid – more ahead’, Policy Brief, 15 July.
Hogan, W. W. (1993), ‘Markets in real electric networks require reactive prices’, Energy Journal, 14
(3), 171–200.
Kahn, E. and R. Baldick (1994), ‘Reactive power is a cheap constraint’, Energy Journal, 15 (4), 191–201.
Netbeheer Nederland (2019), ‘DCC Compliance Verification. Transmission Connected Demand
Facilities, Transmission Connected Distribution Facilities and Distribution Systems. Version 1.0.
Valid from 18 August 2019’.
UCTE (2007), ‘Final Report – System Disturbance on 4 November 2006’.
 How to organize system operation and connection requirements? 125

6A.1 ANNEX: TYPES OF GRID USERS IN THE RfG AND DC NC

Types of Grid Users in the RfG NC

The RfG NC is applicable to power-generating modules, which means either synchronous

power-generating modules or power park modules. An offshore power park module is a type
of power park module that is specified on the basis of its location. According to the definitions
in Article 2 of the RfG NC,

• A power-generating module (PGM) means ‘either a synchronous power-generating

module or a power park module.’
• A synchronous power-generating module (SPGM) means ‘an indivisible set of instal-
lations which can generate electrical energy such that the frequency of the generated
voltage, the generator speed and the frequency of network voltage are in a constant ratio
and thus in synchronism.’
• A power park module (PPM) means ‘a unit or an ensemble of units generating electricity,
which is either non-synchronously connected to the network or connected through power
electronics, and that also has a single connection point to a transmission system, distribu-
tion system including closed distribution system or HVDC system.’
• An offshore power park module (OPPM) means ‘a power park module located offshore
with an offshore connection point.’

A conventional power plant falls under the definition of an SPGM. An asynchronously con-
nected wind power plant is treated as a non-synchronously connected PPM. Solar photovoltaic
or electricity storage devices connected through an inverter are treated as PPMs connected
through power electronics. An AC-connected offshore wind power plant is treated as an
OPPM. Note, however, that an AC-connected onshore wind power plant is treated as a PPM.
Note also that an offshore wind power plant that is connected through an HVDC system is
treated as a DC-connected PPM in the HVDC NC.

Types of Grid Users in the DC NC

The DC NC is applicable to what are defined in its Article 2 as

• Transmission-connected demand facilities, which means ‘a demand facility which has

a connection point to a transmission system’, with a demand facility being defined as
‘a facility which consumes electrical energy and is connected at one or more connection
points to the transmission or distribution system. A distribution system and/or auxiliary
supplies of a power generating module do not constitute a demand facility.’
• Transmission-connected distribution facilities, which means ‘a distribution system
connection or the electrical plant and equipment used at the connection to the transmission
system.’
• Distribution systems, including closed distribution systems, with closed distribution
systems being defined as ‘a distribution system classified pursuant to Article 28 of
Directive 2009/72/EC as a closed distribution system by national regulatory authorities or
by other competent authorities, where so provided by the Member State, which distributes
126 Evolution of electricity markets in Europe

electricity within a geographically confined industrial, commercial or shared services site

and does not supply household customers, without prejudice to incidental use by a small
number of households located within the area served by the system and with employment
or similar associations with the owner of the system.’
• Demand units used by a demand facility or a closed distribution system to provide
demand response services to relevant system operators and relevant TSOs, with
a demand unit being defined as an ‘indivisible set of installations containing equipment
which can be actively controlled by a demand facility owner or by a CDSO [closed
distribution system operator], either individually or commonly as part of demand aggre-
gation through a third party.’ The relevant system operator is defined in the RfG NC
as ‘the transmission system operator or distribution system operator to whose system
a power-generating module, demand facility, distribution system or HVDC system is or
will be connected.’

It is important to note that the DC NC differentiates between ‘system’, ‘facility’ and ‘unit’.
Figure 6A.1 illustrates the use of the terms ‘system’ and ‘facility’ in the DC NC.

Source: Modified from Netbeheer Nederland (2019).

Figure 6A.1 Illustration of the terms ‘demand facility’, ‘distribution facility’ and
‘distribution system’ as used in the DC NC
 How to organize system operation and connection requirements? 127

6A.2 ANNEX: REGULATORY GUIDE

Table 6A.1 Regulatory guide

Section of this chapter, topic and relevant Relevant articles

regulation
Section 6.2
The SO GL formalized TSO coordination Recital 6 states that the rules set out in the SO GL require an institutional
by setting out common requirements for the framework for enhanced coordination between TSOs, including mandatory
establishment of RSCs and for their tasks. participation of TSOs in RSCs. Furthermore, the common requirements for
the establishment of RSCs and for their tasks constitute a first step towards
further regional coordination and integration of system operation.
RSCs deliver five core operational services to Art. 77(3) lists four tasks for RSCs: (a) regional operational security
support national TSO decision-making. coordination; (b) building a common grid model; (c) regional outage
coordination; and (d) regional adequacy assessment. A fifth task, coordinated
capacity calculation, is indirectly mentioned in the CACM GL. Recital 6
of the CACM GL states that ‘capacity calculation for the day-ahead and
intraday should be coordinated at least at regional level.’
RSCs are required to review the relevant Art. 6 of the ER NC states that TSOs shall consult with RSCs to assess the
TSO’s system defence and restoration plans for consistency of measures defined in the system defence and restoration plans
consistency and to provide a technical report to within the entire synchronous area concerned. It states furthermore that within
all the TSOs involved. three months of the submission of the measures by each TSO to the relevant
RSC(s) the RSC(s) shall produce a technical report on the consistency of the
measures based on the criteria set out in paragraph 2 of Art. 6. The report is
to be transmitted without delay to all the TSOs involved, who will in turn
transmit it to the relevant regulatory authorities and ENTSO-E.
The SO GL defines the annual reporting duties Art. 17 states that ‘By 1 March, each regional security coordinator shall
of RSCs to ENTSO-E. prepare an annual report and submit it to ENTSO for Electricity’, providing
information on the tasks it performs.
RCCs are created within the geographical scope Art. 36 of Regulation (EU) 2019/943 states that ‘By 5 January 2020 the
of system operation regions (SORs), a new ENTSO for Electricity shall submit to ACER a proposal specifying which
concept introduced in the CEP. transmission system operators, bidding zones, bidding zone borders, capacity
calculation regions and outage coordination regions are covered by each
of the system operation regions.’ Furthermore, the ‘transmission system
operators of a system operation region shall participate in the regional
coordination centre established in that region.’
All the TSOs in an SOR were required to Art. 35 of Regulation (EU) 2019/943 states that ‘By 5 July 2020, all
forward a proposal for the establishment of transmission system operators of a system operation region shall submit
RCCs for approval to the relevant NRAs by 5 a proposal for the establishment of regional coordination centres to the
July 2020. regulatory authorities concerned …’
Following approval by the NRAs, the RCCs will Art. 35 of Regulation (EU) 2019/943 states that ‘Following approval
replace the RSCs and will enter into operation by by regulatory authorities of the proposal in paragraph 1, the regional
1 July 2022. coordination centres shall replace the regional security coordinators
established pursuant to the system operation guideline … and shall enter into
operation by 1 July 2022.’
128 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
The CEP lists 16 tasks for RCCs. Art. 37 of Regulation (EU) 2019/943 lists the tasks for regional coordination
centres, which are set out in more detail in Annex I of the same regulation.
RCCs shall remain open to extending their Art. 37 of Regulation (EU) 2019/943 states that ‘On the basis of a proposal
portfolio to potential future needs for TSO by the Commission or a Member State, the Committee established by Art.
regional coordination. 68 of Directive (EU) 2019/944 shall issue an opinion on the assignment of
new advisory tasks to regional coordination centres. Where that Committee
issues a favourable opinion on the assignment of new advisory tasks, the
regional coordination centres shall carry out those tasks on the basis of
a proposal developed by the ENTSO for Electricity and approved by ACER
in accordance with the procedure set out in Art. 27.’ In this respect, Art. 68
of Directive (EU) 2019/944 sets out the committee procedure, in which the
‘Commission shall be assisted by a committee [composed of representatives
of the Member States and chaired by a representative of the Commission]
within the meaning of Regulation (EU) No 182/2011.’
RSCs issue coordinated actions to the TSOs. Art. 42 of Regulation (EU) 2019/943 states that ‘Regional coordination
centres shall issue coordinated actions to the transmission system operators
…’
Art. 42 also sets out the procedures described in Subsection 6.2.1 regarding
the implementation of coordinated actions issued by the RSCs.
ENTSO-E can delegate tasks relating to seasonal Art. 6 of the Risk Preparedness Regulation (EU) 2019/941 states that
adequacy assessment and the identification of ENTSO-E ‘may delegate tasks relating to the identification of regional
regional electricity crisis scenarios to RCCs. electricity crisis scenarios to the regional coordination centres.’ Art. 9 states
that ENTSO-E ‘may delegate tasks relating to the [seasonal] adequacy
assessments to regional coordination centres.’
On the basis of these regional electricity crisis Art. 7 of the Risk Preparedness Regulation (EU) 2019/941 states that national
scenarios, Member States shall establish and electricity crisis scenarios shall be identified on the basis of at least the
update their national electricity crisis scenarios, risks listed in Art. 5 of the same regulation for the identification of regional
in principle every four years. electricity crisis scenarios. Furthermore, ‘Member States shall update the
national electricity crisis scenarios every four years, unless circumstances
warrant more frequent updates.’
Section 6.3
The RfG NC applies to new PGMs which Art. 3 states that ‘the connection requirements set out in this Regulation shall
are considered significant, unless otherwise apply to new power-generating modules which are considered significant in
provided. PGMs that are considered significant accordance with Art. 5, unless otherwise provided.’ Art. 5 states that ‘The
need to comply with the requirements of the power-generating modules shall comply with the requirements on the basis
RfG NC according to the voltage level of their of the voltage level of their connection point and their maximum capacity
connection point and their maximum capacity according to the categories set out in paragraph 2’ – types A to D.
(in MW).
 How to organize system operation and connection requirements? 129

Section of this chapter, topic and relevant Relevant articles

regulation
The RfG NC only sets the upper limits (‘upper Art. 5 sets out the limits for thresholds for types B, C and D power-generating
limit minimum capacity thresholds’) for the modules. It states that ‘Proposals for maximum capacity thresholds for types
capacity thresholds used to divide the PGMs B, C and D power-generating modules shall be subject to approval by the
into different types, which differ by synchronous relevant regulatory authority or, where applicable, the Member State.’
area. The final thresholds for the different types
are set at the national level and can be lower
than the maximum threshold, except for PGMs
of type A.
All the RfG NC requirements for lower This is specified in Art. 14 for type B PGMs, Art. 15 for type C PGMs and
categories (e.g. PGM type A) need to be fulfilled Art. 16 for type D PGMs.
by a higher category PGM (e.g. PGM type B).
Furthermore, specific requirements shall apply to Specific requirements are set out in Arts. 17 to 19 for SPGMs of types B to D,
SPGMs of types B to D, PPMs of types B to D Arts. 20 to 22 for PPMs of types B to D, and Arts. 23 to 28 for AC-connected
and AC-connected OPPMs. OPPMs.
All the requirements are minimum requirements, Recital 2 refers to Art. 5 of Directive 2009/72/EC of the Third Package,
which means that PGMs can always have more which requires that ‘Member States or, where Member States have so
enhanced capabilities if this does not impact provided, regulatory authorities ensure, inter alia, that objective and
system security negatively. non-discriminatory technical rules are developed which establish minimum
technical design and operational requirements for the connection to the
system.’
Member States have the obligation to implement Art. 72 states that the SO GL ‘shall enter into force on the twentieth day
the CNCs at the national level no later than three following that of its publication in the Official Journal of the European Union.
years after their publication. Without prejudice to Arts. 4(2)(b), 7, 58, 59, 61 and Title VI, the requirements
of this Regulation shall apply from three years after publication.’
The relevant system operators have two years Art. 7(4) of the RfG NC and Art. 6(4) of the DC NC state that ‘The relevant
to define and submit national specifications for system operator or TSO shall submit a proposal for requirements of general
non-exhaustive requirements for approval by the application, or the methodology used to calculate or establish them, for
competent entity. approval by the competent entity within two years of entry into force of this
Regulation.’
In order to support implementation at the Art. 58 of the RfG NC and Art. 56 of the DC NC state that ‘No later than six
national level and in line with the legal months after the entry into force of this Regulation, the ENTSO for Electricity
requirements for the CNCs, ENTSO-E shall prepare and thereafter every two years provide non-binding written
has an obligation to provide non-binding guidance to its members and other system operators concerning the elements
implementation guidance documents. of this Regulation requiring national decisions. The ENTSO for Electricity
shall publish this guidance on its website.’
For reasons of transparency, ENTSO-E has This is in accordance with Art. 59 of the RfG NC and Art. 57 of the DC NC.
been monitoring the implementation process of
connection network codes across Member States.
The RfG NC gives the relevant TSOs the right Art. 21(2.a) states that ‘the relevant TSO shall have the right to specify that
to specify that PPMs of types C and D shall be [type C] power park modules be capable of providing synthetic inertia during
capable of providing synthetic inertia during very fast frequency deviations.’ Art. 22 states that such requirements are also
very fast frequency deviations to replace the applicable to type D PPMs.
effect of inertia traditionally provided by
SPGMs.
130 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
The RfG NC states that all PGM types shall be Art. 13(1.b) states that ‘With regard to the rate of change of frequency
capable of staying connected to the network and withstand capability, a power-generating module [of type A] shall be
operating at RoCoFs up to a value specified by capable of staying connected to the network and operate at rates of change of
the relevant TSO. frequency up to a value specified by the relevant TSO, unless disconnection
was triggered by rate-of-change-of-frequency-type loss of mains protection.’
As mentioned previously, this is also applicable to PGMs of types B to D.
The RfG NC and the SO GL address ramping Art. 15(6.e) of the RfG NC states that ‘the relevant system operator shall
constraints. specify, in coordination with the relevant TSO, minimum and maximum limits
on rates of change of active power output (ramping limits) in both an up and
down direction of change of active power output for a power-generating
module, taking into consideration the specific characteristics of prime mover
technology.’ Art. 137(4.a) of the SO GL states that all TSOs in an LFC block
shall have the right to determine obligations regarding ramping periods and/
or maximum ramping rates for PGMs and/or demand units in the LFC block
operational agreement.
The DC NC states that demand units offering Art. 28(2.k) states that ‘Demand units with demand response active
demand response shall have the withstand power control, demand response reactive power control, or demand
capability to not disconnect from the system response transmission constraint management shall comply with the
due to RoCoFs up to a value specified by the following requirements, either individually or, where it is not part of
relevant TSO. a transmission-connected demand facility, collectively as part of demand
aggregation through a third party … (k) have the withstand capability to
not disconnect from the system due to the rate-of-change-of-frequency up to
a value specified by the relevant TSO.’
The RfG NC and the DC NC set out Art. 13(1.b) of the RfG NC specifies that a PGM of type A ‘shall be capable
frequency ranges within which PGMs, of remaining connected to the network and operate within the frequency
transmission-connected demand facilities, ranges and time periods’ specified in that article by synchronous area. Art.
transmission-connected distribution facilities, 12 of the DC NC states that ‘Transmission-connected demand facilities,
distribution systems and demand units providing transmission-connected distribution facilities and distribution systems
demand response services should be able to shall be capable of remaining connected to the network and operating at
withstand frequency disturbances for a certain the frequency ranges and time periods specified in Annex I.’ Art. 29(2.a)
period of time without disconnecting from the states that demand units with demand response system frequency control
grid. shallbe able to operate across frequency ranges specified in Art. 12 ‘either
individually or, where it is not part of a transmission-connected demand
facility, collectively as part of demand aggregation through a third party.’
Annex I of the DC NC includes a table with the frequency ranges and time
periods specified by synchronous area.
 How to organize system operation and connection requirements? 131

Section of this chapter, topic and relevant Relevant articles

regulation
The RfG NC foresees that the relevant system Art. 13(1.a.ii) states that ‘the relevant system operator, in coordination with
operator, in coordination with the relevant TSO the relevant TSO and the power-generating facility owner may agree on
and the owner of the PGM, may agree on wider wider frequency ranges, longer minimum times for operation or specific
frequency ranges, longer minimum times for requirements for combined frequency and voltage deviations to ensure
operation or specific requirements for combined the best use of the technical capabilities of a power-generating module, if
frequency and voltage deviations. The DC NC it is required to preserve or to restore system security.’ Art. 12(2) of the
states that the owner of a transmission-connected DC NC states that ‘The transmission-connected demand facility owner
demand facility or the DSO may agree with or the DSO may agree with the relevant TSO on wider frequency ranges
the relevant TSO on wider frequency ranges or or longer minimum times for operation. If wider frequency ranges or
minimum times for operation. longer minimum times for operation are technically feasible, the consent
of the transmission-connected demand facility owner or DSO shall not be
unreasonably withheld.’
A set of requirements relate to the capability Under normal operating conditions, the relevant requirements relate to the
of PGMs and demand units providing demand operation of PGMs of types C and D in frequency sensitive mode as set out
response services to contain or compensate in Art. 15(2.d) and Art. 16(1). Art. 28(2), Art. 29(2) and Art. 30(2) of the
for a frequency drop or rise by regulating the DC NC set out provisions for demand units with demand response active
active power output or input. They include power control, demand response system frequency control and demand
requirements both for operation under normal response very fast active power control respectively.
conditions and for operation in emergency The relevant requirements in emergency situations are related to the capability
situations. of all PGMs to run in limited frequency sensitive mode as specified in Art.
13(2) of the RfG NC. Art. 15(2.c) of the RfG NC adds specific requirements
for types C and D. In addition, Art. 15(2.f) states that PGMs of types C and D
which are capable of acting as a load shall be capable of disconnecting their
load. Finally, Art. 19(1) of the DC NC covers demand disconnection and
reconnection.
The DC NC specifies the reactive power This is in accordance with the provisions in Art. 15. Art. 15(1) states that
requirements for transmission-connected demand ‘Transmission-connected demand facilities and transmission-connected
facilities and transmission-connected distribution distribution systems shall be capable of maintaining their steady-state
systems. operation at their connection point within a reactive power range specified by
the relevant TSO’ in accordance with conditions specified in this article.
The RfG NC and DC NC set out voltage Art. 16(2) of the RfG NC sets out requirements with regard to voltage
ranges within which PGMs of type D, ranges for PGMs of type D. Art. 13(1) of the DC NC states that
transmission-connected demand facilities, ‘Transmission-connected demand facilities, transmission-connected
transmission-connected distribution facilities distribution facilities and transmission-connected distribution systems
and transmission-connected distribution systems shall be capable of remaining connected to the network and operating at
should remain connected to the grid for certain the voltage ranges and time periods specified in Annex II.’ Annex II of the
time periods. DC NC includes tables with voltage ranges and time periods specified by
synchronous area.
132 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
A set of requirements relate to the reactive Under normal operating conditions, the relevant requirements are related
power capabilities of PGMs of types B to D to the right of system operators to specify reactive power capabilities of
and demand units providing demand response SPGMs as set out in Art. 17(2.a) and reactive power capabilities at maximum
services to contain or compensate for voltage and below maximum capacity as set out in Art. 18(2.a–c) and Art. 19(1) of
deviations from the reference value. the RfG NC. Art. 17(2.b), Art. 18(1) and Art. 19(2) of the RfG NC set out
provisions for PGMs to be equipped with a voltage control system. Arts.
20(2.a), 21(3.a–c) and 22 set out requirements for reactive power capabilities
of PPMs. Art. 21(3.d) and Art. 22 define reactive power control modes for
PGMs of types C and D. Art. 28(2) of the DC NC includes provisions for
demand units providing reactive power control. Art. 15(3–4) states that the
relevant TSO may require a transmission-connected distribution system to
actively control the exchange of reactive power at the connection point.
The relevant requirements in emergency situations are related to the capability
of PGMs to automatically disconnect as set out in Arts. 15(3) and 16(2.c) of
the RfG NC. Art. 13(6) and Art. 19(2) of the DC NC set out requirements
for the automatic disconnection of a transmission-connected demand facility,
a transmission-connected distribution facility and a transmission-connected
distribution system, if required by the relevant TSO. Art. 28(3) of the DC NC
sets out requirements for demand units providing demand response services
related to voltage control with disconnection and reconnection.
The RfG NC sets out requirements for PGMs of This is in accordance with Arts. 14(3), 15(1) and (4) and 16(1) and (3) for
types B to D to remain connected to the network PGMs of types B to D.
and operate through periods of low voltage at the
connection point caused by secured faults.
SPGMs of types B to D are required to Art. 17 of the RfG NC states that ‘type B synchronous power-generating
contribute to minimizing the spread of a voltage modules shall be capable of providing post-fault active power recovery. The
dip by quickly recovering their active power relevant TSO shall specify the magnitude and time for active power recovery.’
output following a voltage disturbance. Similar Art. 18(1) and Art. 19(1) cover the applicability for SPGMs of types C and D
specifications exist for PPMs of types B to D. respectively. Regarding PPMs, Art. 20(3) states that ‘the relevant TSO shall
specify the post-fault active power recovery that the power park module is
capable of providing’ and related parameters.
The DC NC sets out requirements for Art. 14 sets out short-circuit requirements. Art. 14(1) states that ‘Based
transmission-connected demand facilities and on the rated short-circuit withstand capability of its transmission network
transmission-connected distribution systems to elements, the relevant TSO shall specify the maximum short-circuit current at
withstand high short-circuit currents to avoid the connection point that the transmission-connected demand facility or the
a potential cascading disconnection of grid users. transmission-connected distribution system shall be capable of withstanding.’
The RfG NC does not mandate black start Art. 15(5.a.i) states that ‘black start capability is not mandatory [for type C
capability for PGMs but refers to the right of PGMs] without prejudice to the Member State’s rights to introduce obligatory
Member States and TSOs to require owners of rules in order to ensure system security.’ Art. 16(1) refers to the applicability
types C and D PGMs to equip their PGMs with for type D PGMs.
a black start capability under certain conditions.
In addition to black start capability, the RfG NC Art. 15(5) and Art. 16(1) specify that PGMs of types C and D shall be capable
also includes system-restoration-related of taking part in island operation if required by the relevant system operator
requirements as regards island operation and in coordination with the TSO. Art. 14(4), Art. 15(1) and (5.c) and Art. 16(1)
reconnection and re-synchronization. set out reconnection and re-synchronization requirements for PGMs of types
B, C and D.
 How to organize system operation and connection requirements? 133

Section of this chapter, topic and relevant Relevant articles

regulation
The CNCs also include requirements related to The RfG NC specifies general system management requirements in Arts.
general system management. 14(5), 15(1) and (6) and 16(1) and (4) for PGMs of types B to D and in Art.
28 for AC-connected OPPMs.
Section 6.4
The scope of application of the DC NC. Art. 3(1) states that ‘the connection requirements set out in this Regulation
shall apply to: (a) new transmission-connected demand facilities; (b) new
transmission-connected distribution facilities; (c) new distribution systems,
including new closed distribution systems; (d) new demand units used by
a demand facility or a closed distribution system to provide demand response
services to relevant system operators and relevant TSOs.’
PGMs and transmission-connected demand Art. 4(2) of the RfG NC specifies when a PGM shall be considered to be
facilities, transmission-connected distribution existing. Art. 4(2) of the DC NC specifies when a transmission-connected
facilities, distribution systems and demand demand facility, a transmission-connected distribution facility, a distribution
units used to provide demand response services system and a demand unit that is, or can be, used by a demand facility or
are considered new if the final and binding a closed distribution system to provide demand response services shall be
contract for the purchase of the main generating considered to be existing.
plant, demand equipment or demand unit is not
finalized by two years after the entry into force
of the relevant regulation.
The RfG and the DC NC can also exceptionally Art. 4(1, 3–5) of the RfG NC and Art. 4(1, 3–5) of the DC NC specify the
apply to existing connections. circumstances under which existing PGMs, transmission-connected demand
facilities, transmission-connected distribution facilities, distribution systems
and demand units that are or can be used by a demand facility or a closed
distribution system to provide demand response services to a relevant system
operator or relevant TSO are subject to the requirements of the respective
regulation.
The RfG NC states that the significance of Recital 9 states that ‘The significance of power-generating modules should
PGMs, and therefore their classification, should be based on their size and their effect on the overall system. Synchronous
be based on their size and their effect on the machines should be classed on the machine size and include all the
overall system. Differences exist between components of a generating facility that normally run indivisibly, such as
synchronously and non-synchronously connected separate alternators driven by the separate gas and steam turbines of a single
generation units. combined-cycle gas turbine installation. For a facility including several
such combined-cycle gas turbine installations, each should be assessed on
its size, and not on the whole capacity of the facility. Non-synchronously
connected power-generating units, where they are collected together to form
an economic unit and where they have a single connection point, should be
assessed on their aggregated capacity.’
The DC NC states that demand connections with This is specified in Art. 3(3) of the DC NC.
more than one demand unit shall be considered
to be one demand unit if they cannot be operated
independently from each other or can reasonably
be considered in a combined manner.
134 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant articles

regulation
The criteria for granting derogations is defined Art. 61(1) of the RfG NC states that ‘Each regulatory authority shall specify,
by the NRA. after consulting relevant system operators and power-generating facility
owners and other stakeholders whom it deems affected by this Regulation,
the criteria for granting derogations pursuant to Articles 62 and 63.’ Similar
provisions are specified in Art. 51 of the DC NC.
Two types of derogations exist depending on Art. 60 of the RfG NC states that ‘Regulatory authorities may, at the request
who files the request. of a power-generating facility owner or prospective owner, relevant system
operator or relevant TSO, grant power-generating facility owners or
prospective owners, relevant system operators or relevant TSOs derogations
from one or more provisions of this Regulation for new and existing power-
generating modules in accordance with Articles 61 to 63.’ Similar powers to
grant derogations are specified in Art. 50 of the DC NC.
All decisions regarding derogations granted or Art. 64 of the RfG NC and Art. 54 of the DC NC state that ‘Regulatory
refused will be notified to ACER and maintained authorities shall maintain a register of all derogations they have granted
in a regularly updated register. or refused and shall provide the Agency with an updated and consolidated
register at least once every six months, a copy of which shall be given to
ENTSO for Electricity.’
ACER and the European Commission both Art. 65(2) of the RfG NC and Art. 54(2) of the DC NC state that ‘The
have the possibility of issuing a reasoned Agency may issue a reasoned recommendation to a regulatory authority to
recommendation to a regulatory authority to revoke a derogation due to a lack of justification. The Commission may issue
revoke a derogation due to a lack of justification. a reasoned recommendation to a regulatory authority or relevant authority of
the Member State to revoke a derogation due to a lack of justification.’
Regulation (EU) 2019/943 foresees a second Art. 59 of Regulation (EU) 2019/943 describes the establishment of
generation of network codes. network codes. Art. 59(1) lists areas where the ‘Commission is empowered
to adopt implementing acts in order to ensure uniform conditions for the
implementation of this Regulation by establishing network codes.’ Art. 59(2)
lists areas where the ‘Commission is empowered to adopt delegated acts
in accordance with Article 68 supplementing this Regulation with regard
to the establishment of network codes.’ Art. 68 refers to the power to adopt
delegated acts that is conferred on the Commission.
The first generation of network codes does not Art. 3 states that the RfG NC shall not apply to ‘(d) storage devices except
cover energy storage, with the exception of for pump-storage power-generating modules in accordance with Article
pump-storage power-generating modules, which 6(2).’ Art. 6(2) refers to the application of the RfG NC and states that
were included in the RfG NC. ‘Pump-storage power-generating modules shall fulfil all the relevant
requirements in both generating and pumping operation mode.’
7. How to ensure adequate investment in power
plants?
Leonardo Meeus with Athir Nouicer

In this chapter, we answer three questions. First, why did some countries introduce a capacity
mechanism? Second, what is the best capacity mechanism? Third, how to limit the (ab)use of
capacity mechanisms?

7.1 WHY DID SOME COUNTRIES INTRODUCE A CAPACITY

MECHANISM?

Capacity mechanisms were introduced to ensure adequate investment in power plants. Today
we talk about resource adequacy rather than generation adequacy because we also have
demand and storage solutions that can compete with power plants in these mechanisms.
However, in some parts of this chapter, we still focus on power plants or generation because
that is how we used to talk about this issue. The mechanisms are called capacity mechanisms
because they provide payment for capacity (MW) in addition to wholesale markets that
remunerate energy (MWh). Electricity markets without capacity mechanisms are sometimes
referred to as energy-only markets. This is misleading because electricity markets without
capacity mechanisms also include capacity components. Retail contracts typically include
capacity components, and the same is true for balancing capacity markets. In what follows,
we focus on the main arguments that have been used in favour of capacity mechanisms. First,
we explain how the economic arguments in favour of capacity mechanisms are evolving from
missing money to missing markets. Second, we argue that the real reason that most countries
go for capacity mechanisms is political rather than economical.
First, the evolution from the missing money argument in favour of capacity mechanisms to
the missing markets argument.1 Electricity demand is relatively inflexible and electric energy
storage capacities are still limited, so prices can go up and down depending on the availability
of renewable energy sources and the demand for electricity. The California crisis in the early
2000s showed that companies can sometimes abuse their market power by driving up electric-
ity prices.2 Around that time, most US electricity markets introduced price caps to limit this
market power abuse. However, this caused some power plants to lose money, typically the
plants that were only needed to cover peak demand. These so-called peakers ran for a limited
number of hours a year. In these hours, they not only had to recover their fuel and other oper-
ating costs, but they also needed to earn a margin that was large enough to cover their capital
costs. If not, they would go out of business. As the price caps were applied equally to all
periods of the year, they also reduced the peakers’ income in times of scarcity. This resulting
loss of income was termed ‘the missing money problem’. Note that the initial focus was on
135
136 Evolution of electricity markets in Europe

price caps, but later it was realized that missing money in wholesale and balancing markets can
also be caused by other interventions. For instance, in Chapter 5 we discussed how capacity
payments to reserve balancing services can distort the scarcity price signal close to real time.
The missing money problem was the main reason why most US wholesale markets introduced
a capacity mechanism at the beginning of this century. The situation in Europe was somewhat
different, and we will come back to it in the next section.
The counter-argument is that capacity mechanisms are not needed. It is simply necessary
to fix the interventions that cause the missing money. For instance, structural measures can
be applied to improve competition so that price caps can be removed and balancing markets
can be reformed. However, there is also a missing market problem in electricity markets. In
Chapter 2, we discussed cross-border hedging in the forward timeframe. The contract length
for transmission rights goes up to a year, while companies typically want to hedge with
longer-duration contracts. In Chapter 6, we referred to technical requirements in connection
agreements that could become markets for system services. Another missing market is the
market for reliability. If there is a shortage in the electricity market, rationing is applied.
Certain districts are disconnected for a few hours, and then other districts for a few hours, so
the pain is spread with what is referred to as a rolling blackout. Unlike a real blackout, this one
is planned and scheduled. In Box 7.1 we discuss a controversy in Belgium around a rationing
plan when the country came close to having to use it. The problem with the rationing approach
is that those who caused the shortage are not held accountable. If retailers knew that their cus-
tomers would be cut off if they do not secure enough supply, they would have more incentives
to enter into long-term contracts with investors in power plants. As long as this is not fixed,
it is another argument in favour of a capacity mechanism to ensure that we have adequate
investment.
Second, the politics of capacity mechanisms.3 Capacity mechanisms have often been
lobbied for by incumbents. These mechanisms are also appealing to risk-averse policymakers
who do not want to be blamed for electricity shortages. In Europe, the number of countries
that have introduced a capacity mechanism increased after the global financial and economic
crisis that started in 2008. Most of the big utilities in Europe had invested substantially in
new gas-fired power plants. Many of these plants did not run in the years following the crisis
because electricity demand stopped growing or even reduced, and renewable energies had
grown faster than expected. In this context, capacity mechanisms became controversial. Some
argued they are simply state aid to national champions that have made the wrong investment
decisions, while others argued they keep alive the gas plants that we desperately need as
backup capacity for renewable energy sources.

BOX 7.1 RATIONING OR LOAD-SHEDDING PLANS AND

CONTROVERSIES

In Belgium, a load-shedding plan led to controversy in the media and in court in 2014. In
the winter of 2014–2015 several nuclear plants were unavailable, so the country risked hav-
ing to use the plan. The general public learned about the plan from the media, which started
to publish maps in six different colours indicating which regions would be cut off first.
Load-shedding plans organize controlled power outages. They make it possible to avoid
a much more serious situation, such as a general blackout of a region. The rationing is usu-
 How to ensure adequate investment in power plants? 137

ally limited in time. Some sites, such as hospitals, are classified as priority and are never
cut. The Belgian plan also avoided city centres and communities with more than 50 000
inhabitants as well as capitals of provinces.
The main controvesy arose in the port of Ghent and the airports of Liège and Charleroi,
which were included in or were close to blue zones – the zones that were first to be cut off.
Ghent demanded that the port area be excluded from the load-shedding plan. The City of
Ghent and the company responsible for managing the port (Havenbedrijf Gent) launched
a court case against the parties involved: the Ministers of Economy and Energy, the trans-
mission system operator (TSO) Elia and the distribution system operator (DSO) IMEWO.
The City asked for €10 million compensation in the case of load shedding. The airports of
Liège and Charleroi also had similar concerns.

Note: Commission de l’économie (2015) summarizes the debate that took place in 2015 regarding
load-shedding plans in Belgium. CREG (2016) presents the modifications in the load-shedding plans that were
decided on following the incident.

7.2 WHAT IS THE BEST CAPACITY MECHANISM?

There are many ways to implement a capacity mechanism, and the relative popularity of the
different options clearly evolved with experience. In what follows, we discuss the experiences
of Spain, Sweden, Great Britain and Ireland, which together nicely illustrate what Europeans
think about capacity mechanisms. We present a taxonomy of capacity mechanisms in Annex
7A.1.
First is the case of Spain, which has been well-documented and commented on by our
colleagues.4 Spain introduced a capacity mechanism as early as 1997. The capacity price was
administratively set by the regulator and the government. In 2007, the mechanism included
two components: investment incentives and availability payments. The investment subsidy
was then €28 000/MW/year. In 2012, the price was reduced from €28 000 to €23 000/MW/
year. In 2013, it was further reduced to €10 000, but the period over which the capacity pay-
ments applied was extended from 10 years to 20 years. The investment incentives were only
for generation capacity installed between 1998 and 2016. In 2018, availability payments for
capacity rewarding combined cycle gas turbines (CCGTs), natural gas, fuel-oil, coal-fired and
hydropower generators were suppressed. The Spanish government’s reason was to be in line
with the Clean Energy Package and the energy transition objectives. This Spanish mechanism
is called capacity payments. A few more countries like Ireland, Italy and Greece also tried out
this mechanism, but it is being abandoned by all of them in favour of mechanisms that set the
price in a market for capacity.
Second is the case of Sweden. Since 2004, it has been possible for power plants to be strate-
gically reserved in Sweden. There is an administrative decision to reserve a certain amount of
capacity, and the price of that capacity is set in an auction. Power plants that would otherwise
be taken offline can then stay on the system and be used in times of shortages. This mechanism
became the most popular mechanism in Europe, even though it received much criticism. The
fear was that governments or regulators would be tempted to use their strategic reserves to
reduce price peaks. This would then act as a de facto price cap in wholesale markets and cause
a missing money problem for the other power plants that did not receive a capacity payment
138 Evolution of electricity markets in Europe

from the strategic reserve fund. In Sweden, reserve activation only happens if there is a curtail-
ment situation in Sweden or Finland. TSOs in both countries decide together which resources
to use in the case of shortages. Note that today a large part of the Swedish strategic reserve is
demand response from load-heavy industry.
Third is the case of Great Britain. Since 2014, Great Britain has had a US-style capacity
auction, often referred to as a capacity market. There is an administrative decision to procure
a certain amount of capacity, and the price of that capacity is set in an auction. The capacity
market applies to all the power plants or resources able to provide capacity. The problem with
such a mechanism is that it distorts the wholesale market. The wholesale market is a European
market in which power plants from different countries compete, as was explained in Chapter
2, but some of them are now supported by national capacity mechanisms while others are not.
Within a national market with a capacity mechanism, the mechanism also distorts the level
playing field between different players and technologies. In the past these mechanisms focused
on supporting power plants, while today supply solutions are wanted to compete with demand
and energy storage solutions. It has proven to be very difficult to design capacity mechanisms
that are technology-neutral. The so-called Tempus case illustrates the controversies around
these mechanisms (Box 7.2).

BOX 7.2 THE TEMPUS CASE

Tempus, a demand response provider, argued that the capacity market scheme in Great
Britain discriminated against demand response. The mechanism had been approved by the
European Commission, which was challenged by Tempus. In 2018 the General Court of
the European Union issued a judgment in case T-793/14 ‘Tempus Energy Ltd and Tempus
Energy Technology Ltd v Commission’ overruling the Commission’s decision to approve
the state aid scheme establishing a capacity market in the United Kingdom (UK). The UK
government suspended the capacity market to comply with the judgment. This led to an
immediate stopping of the existing capacity payments and the cancelling of 2019 auctions.
The European Commission appealed against the decision and conducted a state aid in-
vestigation. The UK government supported the Commission’s appeal. After an in-depth
investigation, in October 2019 the Commission again approved the mechanism. It con-
firmed that it complied with the EU state aid rules and that it did not distort competition in
the single market. The investigation did not find evidence that the capacity market scheme
would put demand response providers at a disadvantage. At the same time, the UK gov-
ernment committed to improving the scheme, considering recent market and regulatory
developments and other issues identified in the UK government’s 2019 five-year review
of the capacity market. This includes, for instance, revisiting the minimum capacity size to
participate in the auction and the rules on participation by new types of players. Tempus can
still challenge the re-approval, and other related proceedings are still ongoing.

Note: Before the Clean Energy Package was adopted, the European Commission instruments to control the
implementation of capacity mechanisms were the EU state aid rules. Our colleagues Hancher et al. (2015) are
experts on state aid and provide a detailed discussion of experiences with capacity mechanisms.
 How to ensure adequate investment in power plants? 139

Fourth is the case of Ireland. Ireland started with Spanish-style capacity payments but became
the first country in Europe to introduce reliability options. This is a mechanism that has been
advocated by academics as superior to capacity markets because it pays for the availability of
capacity at times of shortages rather than paying for capacity. Reliability options are based on
the concept of a call options contract, and capacity providers enter into an option contract with
a counterparty (a TSO or a large consumer or supplier). This contract offers the counterparty
the option of procuring electricity at a predetermined strike price. In return for the premium
paid – the price of the option contract – the counterparty gets insurance against high prices.
This premium then replaces the capacity remuneration. The mechanism incentivizes power
plants or other players to have resources available when prices go up. If they do not, they will
have to buy energy at the high prices in the market and sell it at the lower strike price that has
been agreed when entering the insurance scheme. In addition, regulators may apply further
safeguards with a penalty for non-compliance and/or by requiring a certain level of physical
capacity for all options sold. By design, this mechanism is less distorting for the wholesale
market than capacity markets. However, when we looked into the implementation details, we
found that many parameters are still set administratively. These include who can offer which
amount of reliability options, the length of the contract and many other factors that are known
to favour some solutions over others.5
In other words, there is no best design for capacity mechanisms. If a country is fully
convinced its electricity market cannot survive without a capacity mechanism, reliability
options are the most elegant design option, but it will need to keep a close eye on the imple-
mentation details. If it is not convinced but political reality pushes it towards an intervention
to safeguard the availability of capacity in the system, strategic reserves will be its preferred
option. They are easier to implement and the Clean Energy Package includes provisions that
remedy the main risks associated with strategic reserves, as we will discuss in the next section.

7.3 HOW TO LIMIT THE (AB)USE OF CAPACITY

MECHANISMS?

The Clean Energy Package goes a long way towards limiting the (ab)use of capacity mecha-
nisms. Regulation (EU) 2019/943 introduced two steps to check if capacity mechanisms are
really needed and also includes provisions to guide the design of these mechanisms.

7.3.1 Is a Capacity Mechanism Really Needed?

Here, we introduce the European resource adequacy assessment and the national implemen-
tation plan for market reforms, both of which will be used to check whether a certain country
needs a capacity mechanism.
First, the European resource adequacy assessment. Following Regulation (EU) 2019/943,
this assessment will be used to check if there is a need for concern. The assessment is to be
carried out every year by ENTSO-E based on data provided by national TSOs and cover
a period of ten years. A European assessment can avoid over-reaction if a certain country has
an issue which can be solved by its neighbours. It can also avoid under-reaction if a number
of countries are counting on the same neighbour and/or over-estimate the capacity of their
neighbours to help in certain situations. Member States can also continue to do national assess-
140 Evolution of electricity markets in Europe

ments, which they can use to complement the European assessment with additional sensitivity
analysis. If this leads to disputes, a process has been foreseen with a role for the Agency for
the Cooperation of Energy Regulators (ACER).
Regulation (EU) 2019/943 introduced a process to come up with a methodology that will be
used in these assessments. The role of the Member States or a competent authority designated
by them is to define a reliability standard. The reliability standard is the level of risk countries
want to take to face shortages, which is expressed as the ‘expected energy not served’ (EENS),
or the ‘loss of load expectation’ (LOLE). The role of the European Network of Transmission
System Operators for Electricity (ENTSO-E), with oversight from ACER, is to integrate this
standard into a methodology that also considers the cost of new entry of generation, or demand
response, and the cost people face when they do not have electricity, which is expressed as the
value of lost load (VoLL). Scenarios with demand and supply projections will also need to be
agreed upon to be able to perform this assessment. In December 2019, ENTSO-E launched
two public consultations to collect inputs from stakeholders on the proposal for the EU
resource adequacy assessment methodology and on the VoLL, the cost of a new entry and the
reliability standard calculation methodology.
Second, the national implementation plan for market reforms. As discussed in the previ-
ous sections, much can be done to reduce the need for capacity mechanisms. Price caps can
be removed in wholesale markets and retail markets, more investment can be made in the
transmission network and balancing markets can be reformed, as was discussed in Chapter
5. Following Regulation (EU) 2019/943, Member States with identified resource adequacy
concerns need to develop and publish an implementation plan with a timeline for adopting
measures to eliminate any identified market distortions. The plans will be reviewed by the
European Commission, which shall issue an opinion within four months. Member States also
have to monitor the application of their plans and publish the results of the monitoring in an
annual progress report.
Regulation (EU) 2019/943 refers to capacity mechanisms as measures of last resort to
eliminate residual resource adequacy concerns. Member States are to carry out a study of the
possible effects of capacity mechanisms on neighbouring countries before they are imple-
mented. They are also to prioritize assessment of the potential of a strategy reserve mechanism
to address residual resource adequacy concerns. Only if strategic reserves cannot address them
may Member States implement other types of capacity mechanisms.
Capacity mechanisms are also to be temporary and to be approved by the European
Commission for a maximum of ten years. Then they are to be phased out, or the amount of
committed capacity is to be reduced according to the national market reform implementation
plan. Member States are to continue the application of the implementation plan after the intro-
duction of the capacity mechanism. They need to include a provision allowing for an efficient
administrative phase-out in the case that no new contracts are concluded for three consecutive
years. Member States are to review capacity mechanisms that are already in place. If the
resource adequacy assessments have not identified concerns, no new contracts under these
mechanisms are to be concluded.
 How to ensure adequate investment in power plants? 141

7.3.2 Design Principles for Capacity Mechanisms

Regulation (EU) 2019/943 provides guidance to Member States regarding the design of capac-
ity mechanisms. Below, we discuss the design principles that apply to all mechanisms, and
specific ones that apply to strategic reserves.
First is technology neutrality. Capacity mechanisms need to be open to participation by
all capable resources, including demand and storage solutions and decentralized energy
resources. Resources should be selected through a transparent non-discriminatory competitive
process. The mechanisms must also provide incentives for participants to be available in times
of system stress when they are most needed. Appropriate penalties should be applied for pro-
viders who are not available during such times.
Second are strategic reserves. Strategic reserves are only to be dispatched in the case that
TSOs are likely to exhaust their balancing resources. During imbalance settlement periods
in which resources in the strategic reserve are dispatched, the imbalances in the market are
to be settled at least at the VoLL or at a higher value than the intraday technical price limit,
whichever is higher. Following the dispatch, the strategic reserve’s output is to be attributed to
balance responsible parties (BRPs) through the imbalance settlement mechanism. In addition,
the resources participating in the strategic reserve mechanism shall not receive any remuner-
ation from wholesale or balancing markets. They are to be held outside the market at least
during the contractual period.
Third are CO2 emission limits. Capacity mechanisms cannot be abused to subsidize the
most polluting coal power plants. A grandfathering clause was introduced for capacity mecha-
nism contracts that were concluded before 31 December 2019. Since 4 July 2019, a CO2 emis-
sion limit of 550 g of CO2 of fossil fuel origin per kWh has been applied for new generation
capacity. Additionally, from 1 July 2025, an emission limit of 350 kg of CO2 of fossil fuel
origin on average per year per installed kWe for existing capacity is applied, that is, generation
capacity that started commercial production before 4 July 2019. In December 2019, ACER
published an opinion providing technical guidance related to the calculation of the values of
CO2 emission limits for capacity mechanisms. It clarifies the scope of application of the emis-
sion limits and sets out the calculation formulae and the related specifications.
Fourth is cross-border participation. Capacity mechanisms other than strategic reserves
must be open to explicit cross-border participation to limit distortions to cross-border trade and
competition and to provide incentives for interconnection investment to ensure the EU security
of electricity supply at the least cost. Cross-border participation in strategic reserves is to be
open where technically feasible. With ACER oversight, ENTSO-E needs to develop a meth-
odology to calculate the maximum entry capacity for cross-border participation. TSOs are then
to set the maximum entry capacity for foreign capacity on the basis of a recommendation by
the regional coordination centre. Many more details will be worked out to enable cross-border
participation in capacity mechanisms.

7.4 CONCLUSION

In this chapter on how to ensure adequate investment in power plants, we have answered three
questions. First, why did some countries introduce a capacity mechanism to ensure adequate
investment in power plants? The arguments in favour of capacity mechanisms are evolving
142 Evolution of electricity markets in Europe

from missing money to missing markets. The missing money problem was first identified in
the US, where price caps had led to a reduction in generation unit incomes, especially those
of peakers. Missing markets include longer-term hedging across borders and the market for
reliability. We also showed that the real reason behind capacity mechanisms is often political
rather than economical.
Second, what is the best capacity mechanism? We discussed the experiences of Spain with
capacity payments, Sweden with strategic reserves, GB with capacity markets and Ireland
with reliability options. We cannot say that there is an absolute winner. Nevertheless, there is
a trend towards strategic reserves, which is reflected in the Clean Energy Package framework
for capacity mechanisms.
Third, how to limit the (ab)use of capacity mechanisms? We presented the new regula-
tory framework introduced in the Clean Energy Package. It includes EU resource adequacy
assessment and national implementation plans for market reforms. Member States need to
check whether it is sufficient to introduce strategic reserves, and detailed design principles
for these reserves are provided. The new framework also encompasses other design principles
for capacity mechanisms, such as technology neutrality, CO2 emission limits and rules on
cross-border participation.

NOTES
1. Shanker (2003) used the missing money concept in his reaction to the standard market design pro-
posals by the Federal Energy Regulatory Commission (FERC) in the US. Cramton and Stoft (2006)
have also been influential in their study of the issue. Joskow (2008) went a step further and argued
that the missing money is not only caused by price caps but there are other market interventions that
have the same effect. Examples include actions by system operators to ensure system reliability.
Newbery (1989) discusses the missing market problem when revenue is adequate but not perceived,
that is, the commodity being sold is considered a public good. Bhagwat et al. (2016) present a survey
of US experts regarding capacity mechanisms. The answers are in favour of energy-only markets.
Cretì and Fontini (2019) highlight the pros and cons of the different capacity mechanisms. They
provide an alternative tool to capacity mechanisms: the operating reserve demand curve, which
complements energy-only markets. This tool is also discussed in Hogan (2013), who argues that it
can complement both energy-only markets and markets with capacity mechanisms. We referred to
this approach in Chapter 5.
2. The California crisis in 2000–2001 is one of the most iconic electricity market crises. Several
academics from the US have analysed the California crisis including Borenstein (2002), Borenstein
et al. (2003, 2008), Joskow (2001, 2008) and Wolak (2003). In 1998, the state had introduced
competitive wholesale and retail markets for electricity. It was the first fully decentralized market
arrangement, with a power exchange (CALPX) and an independent system operator (CAISO),
in the US. CAISO operated the transmission networks owned by the three major utilities and ran
various energy-balancing, ancillary services and congestion management markets, while CALPX
ran both a voluntary day-ahead and hourly hour-ahead public wholesale markets for energy. Note
that the three largest utilities had a (largely unhedged) obligation to trade on the day-ahead and
real-time markets operated by the PX and the ISO. Wholesale market prices were deregulated, while
retail prices were fixed for up to four years. The reforms assumed that wholesale prices would be
lower than the administrative cap set for the retail price. As it turned out, markets appeared to be
quite competitive in periods with low and moderate demand. However, demand had been increasing
rapidly in the previous years, while there had been little new investment. As a result, in periods of
high demand a combination of tight supplies and inelastic demand created opportunities for individ-
ual generators to exercise market power. Between May 2000 and June 2001, California experienced
an explosion in wholesale power prices, followed by supply shortages and rolling blackouts with
 How to ensure adequate investment in power plants? 143

physical rationing by the ISO. In January 2001, CALPX declared itself bankrupt after it had been
unable to implement mitigation measures imposed by the regulator. California responded to the
crisis with costly long-term contracts of up to 20 years negotiated by the state, long-term obligations
for procurement, a freeze on retail competition and, overall, by taking some step backs from a fully
decentralized market model. The California crisis even made it into the documentary on the Enron
bankruptcy The Smartest Guys in the Room.
3. Thomas-Olivier Léautier (2019) argues that capacity mechanisms owe more to the political
economy than to microeconomics. He argues that elected officials prefer a zero-blackout criterion
irrespective of the costs incurred. He also argues that most system operators, regulators and govern-
ment employees are in favour of an increase in their scope of intervention through the design and
monitoring of capacity mechanisms. These employees are the first to be blamed by politicians in the
case of blackouts.
4. In Batlle et al. (2007) and Batlle and Pérez-Arriaga (2008) our colleagues share their views on the
Spanish and international experiences with capacity mechanisms.
5. The concept of reliability option was first outlined by Pérez-Arriaga (1999). Vázquez et al. (2002)
discuss the experience of Colombia with this mechanism. Oren (2005) also promotes the concept.
In Bhagwat and Meeus (2019) we discuss the implementation of reliability options in Ireland and
Italy.

REFERENCES
ACER (2013), ‘Capacity Remuneration Mechanisms and the Internal Market for Electricity’.
Batlle, C. and I. J. Pérez-Arriaga (2008), ‘Design criteria for implementing a capacity mechanism in
deregulated electricity markets’, Utilities Policy, 16 (3), 184–93.
Batlle, C., C. Vázquez, M. Rivier and I. J. Pérez-Arriaga (2007), ‘Enhancing power supply adequacy
in Spain: Migrating from capacity payments to reliability options’, Energy Policy, 35 (9), 4545–54.
Bhagwat, P. and L. Meeus (2019), ‘Reliability options: Can they deliver on their promises?’, Electricity
Journal, 32 (10).
Bhagwat, P. C., L. J. de Vries and B. F. Hobbs (2016), ‘Expert survey on capacity markets in the US:
Lessons for the EU’, Utilities Policy, 38, 11–17.
Borenstein, S. (2002), ‘The trouble with electricity markets: Understanding California’s restructuring
disaster’, Journal of Economic Perspectives, 16 (1), 191–211.
Borenstein, S., J. Bushnell, C. R. Knittel and C. Wolfram (2008), ‘Inefficiencies and market power in
financial arbitrage: A study of California’s electricity markets’, Journal of Industrial Economics, 56
(2), 347–78.
Borenstein, S., J. Bushnell and F. A. Wolak (2003), ‘Measuring market inefficiencies in California’s
restructured wholesale electricity market’, American Economic Review, 92 (5), 1376–405.
Commission de l’économie (2015), ‘Débat Sur Le Plan de Délestage’ [‘Debate on the Shedding Plan’],
accessed at https://​www​.dekamer​.be/​doc/​CCRI/​pdf/​54/​ic044x​.pdf.
Cramton, P. and S. Stoft (2006), ‘The Convergence of Market Designs for Adequate Generating
Capacity’, Report for California Electricity Oversight Board, accessed at https://​doi​.org/​10​.1007/​
s13398​-014​-0173​-7​.2.
CREG (2016), ‘Rapport Annuel 2015’ [Annual Report 2015’].
Cretì, A. and F. Fontini (2019), ‘Investing in power generation’, in Economics of Electricity: Markets,
Competitions and Rules, Cambridge: Cambridge University Press, pp. 259–98.
EP (2017), ‘Capacity Mechanisms for Electricity’, accessed at http://​www​.europarl​.europa​.eu/​RegData/​
etudes/​BRIE/​2017/​603949/​EPRS​_BRI(2017)603949​_EN​.pdf.
Hancher, L., A. de Hauteclocque and M. Sadowska (ed.) (2015), Capacity Mechanisms in the EU Energy
Market: Law, Policy, and Economics, Oxford: Oxford University Press.
Hogan, W. W. (2013), ‘Electricity scarcity pricing through operating reserves’, Economics of Energy and
Environmental Policy, 2 (2).
Joskow, P. (2001), ‘California’s electricity crisis’, Oxford Review of Economic Policy, 17 (3), 365–88.
Joskow, P. (2008), ‘Lessons learned from electricity market liberalization’, Energy Journal, 29 (Special
Issue #2), 9–42.
144 Evolution of electricity markets in Europe

Léautier, Thomas-Olivier (2019), ‘The highly visible hand: Capacity mechanism’, in Imperfect Markets
and Imperfect Regulation, Cambridge, MA: MIT Press, pp. 325–59.
Newbery, D. M. (1989), ‘Missing markets: Consequences and remedies’, in F. H. Hahn (ed.), The
Economics of Missing Markets, Information, and Games, Oxford: Clarendon Press, pp. 211–42.
Oren, S. S. (2005), ‘Generation adequacy via call options obligations: Safe passage to the promised land’,
Electricity Journal, 18 (9), 28–42.
Pérez-Arriaga, I. J. (1999), ‘Reliability and generation adequacy’, in W. S. Read, W. K. Newman, I.
J. Pérez-Arriaga, H. Rudnick, M. R. Gent and A. J. Roman (1999), ‘Reliability in the new market
structure (Part 1) ’, IEEE Power Engineering Review, 19 (12), 4–5.
Shanker, R. (2003), ‘Comments on Standard Market Design: Resource Adequacy Requirement. Federal
Energy Regulatory Commission’, accessed at http://​elibrary​.ferc​.gov/​idmws/​common/​opennat​.asp​
?fileID​=​9619272.
Vázquez, C., M. Rivier and I. J. Pérez-Arriaga (2002), ‘A market approach to long-term security of
supply’, IEEE Transactions on Power Systems, 17 (2), 349–357.
Wolak, F. A. (2003), ‘Diagnosing the California electricity crisis’, Electricity Journal, 16 (7), 11–37.
 How to ensure adequate investment in power plants? 145

7A.1 ANNEX: TAXONOMY OF CAPACITY MECHANISMS

Table 7A.1 Taxonomy of capacity mechanisms

Type Taxonomy of capacity mechanisms based on EP (2017) and ACER (2013)

Strategic reserve: A central agency (transmission system operator or a government agency) decides on
the amount of capacity needed to make up any shortfall in the market a few years in advance. The level
of payment for the contracted capacity (strategic reserve) is set through a competitive tendering process.
The contracted power plants cannot participate in the electricity market and are only activated in the case
Targeted

of extreme conditions. The 2016 Commission Sector Inquiry highlighted a strategic reserve as the most
appropriate mechanism for circumstances where temporary or local adequacy concerns are identified.
Capacity obligation: Also called capacity requirement, this is an obligation on suppliers or large consumers
to contract with generators for a certain level of capacity related to their self-assessed future consumption or
supply (e.g. three years ahead) plus a reserve margin that is decided on by an independent body. If not
enough capacity is contracted, the supplier or the consumer will pay a buy-out price/fine. The price for
capacity is determined in a decentralized way through the contracts. This model can also include a market for
exchangeable obligations (secondary market).
Capacity auction: The capacity volume to be auctioned is decided centrally (by the TSO or regulator) and
totally a few years in advance. The price is determined by auction and is paid to all resources (existing and
new) clearing the auction. Capacity providers bid to receive a payment that reflects the cost of building new
capacity. The new capacity participates in the energy-only market.
Reliability options: This mechanism is based on a forward auction (e.g. three years ahead). A capacity
provider enters into an option contract with a counterparty (a TSO or a large consumer or supplier). The
Volume-based

contract offers the counterparty the option to procure electricity at a predetermined strike price. The capacity
Market-wide

provider must be available to the system operator for dispatch above the strike price. In the 2016 Commission
Sector Inquiry, reliability options were highlighted as the most appropriate mechanism where long-term
adequacy concerns are identified.
Capacity payments: This is a price-based mechanism. It pays a fixed amount (set by the regulator) for the
Price-based

capacity available to all generators. The plants receiving capacity payments continue to participate in the
energy-only market. The payment can also be made when the plant does not run, but certain availability
criteria have to be met.
146 Evolution of electricity markets in Europe

7A.2 ANNEX: REGULATORY GUIDE

Table 7A.2 Regulatory guide

Section of this chapter, topic and relevant Relevant article

regulation
Section 7.3.1
Regulation (EU) 2019/943 introduces a binding Art. 20(1) states that ‘Member States shall monitor resource adequacy within
EU resource adequacy assessment. It can be their territory on the basis of the European resource adequacy assessment
complemented with national assessments. referred to in Article 23. For the purpose of complementing the European
resource adequacy assessment, Member States may also carry out national
resource adequacy assessments pursuant to Article 24.’When a resource
adequacy concern is identified in an assessment, Art. 20(2) states that ‘… the
Member State concerned shall identify any regulatory distortions or market
failures that caused or contributed to the emergence of the concern.’
Regulation (EU) 2019/943 sets the coverage Art. 23(1) states that ‘the European resource adequacy assessment shall
and duration of the EU resource adequacy identify resource adequacy concerns by assessing the overall adequacy of the
assessment together with ENTSO-E’s role in electricity system to supply current and projected demands for electricity at
its conduct. Union level, at the level of the Member States, and at the level of individual
bidding zones, where relevant. The European resource adequacy assessment
shall cover each year within a period of 10 years from the date of that
assessment.’Art. 23(2) adds that ‘the European resource adequacy assessment
shall be conducted by the ENTSO for Electricity.’
Regulation (EU) 2019/943 provides Art. 23(3) states that ‘by 5 January 2020, the ENTSO for Electricity shall
a development process for the draft submit to the Electricity Coordination Group set up under Article 1 of
methodology for the EU resource adequacy Commission Decision of 15 November 2012 and ACER a draft methodology
assessment and the principles underlying it. for the European resource adequacy assessment based on the principles
provided for in paragraph 5 of this Article.’
Among the principles underlying the EU adequacy assessment which are to be
ensured by the transparent methodology, Art. 23(5) states that the assessment
is to be carried out at each bidding zone level. It is to be based on appropriate
central reference scenarios of projected demand and supply. In addition, the
methodology must ensure that the assessment ‘appropriately takes account of
the contribution of all resources including existing and future possibilities for
generation, energy storage, sectoral integration, demand response, and import
and export and their contribution to flexible system operation.’ Moreover,
the assessment should apply at least the ‘expected energy not served’ (EENS)
and the ‘loss of load expectation’ (LOLE) indicators, which are referred to in
Art. 25.
Regulation (EU) 2019/943 requires the Art. 23(6) states that ‘by 5 January 2020, the ENTSO for Electricity shall
development of a methodology for VoLL, cost submit to ACER a draft methodology for calculating:
of new entry and reliability standards. (a) the value of lost load;
(b) the cost of new entry for generation, or demand response; and
(c) the reliability standard referred to in Article 25.
The methodology shall be based on transparent, objective and verifiable
criteria.’
 How to ensure adequate investment in power plants? 147

Section of this chapter, topic and relevant Relevant article

regulation
According to Regulation (EU) 2019/943, the Art. 25(1) states that ‘when applying capacity mechanisms Member States
reliability standard aims to reflect the level of shall have a reliability standard in place. A reliability standard shall indicate
the security of supply of the Member States. the necessary level of security of supply of the Member State in a transparent
Regulation (EU) 2019/943 sets principles for manner. In the case of cross-border bidding zones, such reliability standards
the calculation of the reliability standard. shall be established jointly by the relevant authorities.’Art. 25(2) adds that
‘the reliability standard shall be set by the Member State or by a competent
authority designated by the Member State, following a proposal by the
regulatory authority …’In addition the reliability standardis to be‘… calculated
using at least the value of lost load and the cost of new entry over a given
timeframe and shall be expressed as “expected energy not served” and “loss
of load expectation”’(Art. 25(3)).
Regulation (EU) 2019/943 provides a guide for Art. 20(3) lists the principles that are to be taken into account when addressing
the development and publication of national adequacy concerns. It states that ‘Member States with identified resource
implementation plans for market reforms. adequacy concerns shall develop and publish an implementation plan with
a timeline for adopting measures to eliminate any identified regulatory
distortions or market failures as a part of the State aid process. When
addressing resource adequacy concerns, the Member States shall in particular
take into account the principles set out in Article 3 and shall consider:
(a) removing regulatory distortions;
(b) removing price caps in accordance with Article 10;
(c) introducing a shortage pricing function for balancing energy as referred to
in Article 44(3) of Regulation (EU) 2017/2195;
(d) increasing interconnection and internal grid capacity with a view to
reaching at least their interconnection targets as referred in point (d)(1) of
Article 4 of Regulation (EU) 2018/1999;
(e) enabling self-generation, energy storage, demand side measures and
energy efficiency by adopting measures to eliminate any identified regulatory
distortions;
(f) ensuring cost-efficient and market-based procurement of balancing and
ancillary services;
(g) removing regulated prices where required by Article 5 of Directive (EU)
2019/944.’
Regulation (EU) 2019/943 sets out a process for Art. 20(4) states that the Member States concernedshall ‘submit their
the submission and review of implementation implementation plans to the Commission for review.’Then, ‘within four months
plans. of receipt of the implementation plan, the Commission shall issue an opinion
on whether the measures are sufficient to eliminate the regulatory distortions
or market failures that were identified pursuant to paragraph 2, and may invite
the Member States to amend their implementation plans accordingly’ (Art
20(5)). The application of the implementation plans shall be monitored by the
Member States concerned, as is stated in Art. 20(6). An annual report on the
results of the monitoring is to be published by the Member States concerned
and submitted to the Commission, which ‘shall issue an opinion on whether
the implementation plans have been sufficiently implemented and whether the
resource adequacy concern has been resolved’(Art. 20(7)). Finally, Art. 20(8)
states that ‘Member States shall continue to adhere to the implementation plan
after the identified resource adequacy concern has been resolved.’
148 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant article

regulation
Regulation (EU) 2019/943 allows the Art. 21(1) states that ‘to eliminate residual resource adequacy concerns,
implementation of a capacity mechanism as Member States may, as a last resort while implementing the measures referred
a measure of last resort. to in Article 20(3) of this Regulation in accordance with Article 107, 108 and
109 of the TFEU, introduce capacity mechanisms.’
Regulation (EU) 2019/943 requires Member Art. 21(2) states that ‘before introducing capacity mechanisms, the Member
States to conduct a study of the effect on States concerned shall conduct a comprehensive study of the possible effects of
neighbouring Member States. such mechanisms on the neighbouring Member States by consulting at least its
neighbouring Member States to which they have a direct network connection
and the stakeholders of those Member States.’
Regulation (EU) 2019/943 gives priority to Art. 21(3) requires that Member States give priority to capacity mechanisms
strategic reserves implementation. in the form of strategic reserves. It states that ‘Member States shall assess
whether a capacity mechanism in the form of strategic reserve is capable
of addressing the resource adequacy concerns. Where this is not the case,
Member States may implement a different type of capacity mechanism.’
Regulation (EU) 2019/943 includes a provision Art. 21(7) states that ‘when designing capacity mechanisms Member States
on the phasing-out of capacity mechanisms. shall include a provision allowing for an efficient administrative phase-out
of the capacity mechanism where no new contracts are concluded under
paragraph 6 during three consecutive years.’
Regulation (EU) 2019/943 specifies that Art. 21(8) states that capacity mechanisms are to be temporary.It adds that
capacity mechanisms are temporary. ‘they shall be approved by the Commission for no longer than 10 years. They
shall be phased out or the amount of the committed capacities shall be
reduced on the basis of the implementation plans referred to in Article 20.
Member States shall continue to apply the implementation plan after the
introduction of the capacity mechanism.’
Regulation (EU) 2019/943 includes measures Regarding capacity mechanisms already in place, Art. 21(6) says that‘where
on already existing capacity mechanisms. a Member State applies a capacity mechanism, it shall review that capacity
mechanism and shall ensure that no new contracts are concluded under that
mechanism where both the European resource adequacy assessment and
the national resource adequacy assessment, or in the absence of a national
resource adequacy assessment, the European resource adequacy assessment
have not identified a resource adequacy concern or the implementation plan as
referred to in Article 20(3) has not received an opinion by the Commission as
referred to in Article 20(5).’
Section 7.3.2
Regulation (EU) 2019/943 provides general Art. 22(1) enumerates design principles for capacity mechanisms. It states that
design principles for capacity mechanisms. ‘any capacity mechanism shall:
(a) be temporary;
(b) not create undue market distortions and not limit cross-zonal trade;
 How to ensure adequate investment in power plants? 149

Section of this chapter, topic and relevant Relevant article

regulation
(c) not go beyond what is necessary to address the adequacy concerns referred
to in Article 20;
(d) select capacity providers by means of a transparent, non-discriminatory
and competitive process;
(e) provide incentives for capacity providers to be available in times of
expected system stress;
(f) ensure that the remuneration is determined through the competitive
process;
(g) set out the technical conditions for the participation of capacity providers
in advance of the selection process;
(h) be open to participation of all resources that are capable of providing the
required technical performance, including energy storage and demand side
management;
(i) apply appropriate penalties to capacity providers that are not available in
times of system stress.’
Regulation (EU) 2019/943 provides specific Art. 22(2) specifies design principles for strategic reserves. It states that ‘the
design principles for strategic reserves. design of strategic reserves shall meet the following requirements:
(a) where a capacity mechanism has been designed as a strategic reserve,
the resources thereof are to be dispatched only if the transmission system
operators are likely to exhaust their balancing resources to establish an
equilibrium between demand and supply;
(b) during imbalance settlement periods where resources in the strategic
reserve are dispatched, imbalances in the market are to be settled at least at
the value of lost load or at a higher value than the intraday technical price
limit as referred in Article 10(1), whichever is higher;
(c) the output of the strategic reserve following dispatch is to be attributed to
balance responsible parties through the imbalance settlement mechanism;
(d) the resources taking part in the strategic reserve are not to receive
remuneration from the wholesale electricity markets or from the balancing
markets;
(e) the resources in the strategic reserve are to be held outside the market for
at least the duration of the contractual period.
The requirement referred to in point (a) of the first subparagraph shall be
without prejudice to the activation of resources before actual dispatch in
order to respect the ramping constraints and operating requirements of the
resources. The output of the strategic reserve during activation shall not be
attributed to balance groups through wholesale markets and shall not change
their imbalances.’
150 Evolution of electricity markets in Europe

Section of this chapter, topic and relevant Relevant article

regulation
Regulation (EU) 2019/943 provides design For mechanisms other than strategic reserves, Art. 22(3) lists additional design
principles for mechanisms other than strategic principles to those included in Art. 22(1). It states that‘capacity mechanisms
reserves. other than strategic reserves shall:
(a) be constructed so as to ensure that the price paid for availability
automatically tends to zero when the level of capacity supplied is expected to
be adequate to meet the level of capacity demanded;
(b) remunerate the participating resources only for their availability and
ensure that the remuneration does not affect decisions of the capacity provider
on whether or not to generate;
(c) ensure that capacity obligations are transferable between eligible capacity
providers.’
Regulation (EU) 2019/943 sets emission limits Art. 22(4) includes two limits on the CO2 emissions of generating units
for generation units to participate in capacity participating in capacity mechanisms. This is with regard to the start date
mechanisms. It differentiates between existing of their commercial production. It states that ‘capacity mechanisms shall
and new generation capacity and requires incorporate the following requirements regarding CO2 emission limits:
ACER to provide technical guidance on the (a) from 4 July 2019 at the latest, generation capacity that started commercial
emission limit values. production on or after that date and that emits more than 550 g of CO2 of
fossil fuel origin per kWh of electricity shall not be committed or to receive
payments or commitments for future payments under a capacity mechanism;
(b) from 1 July 2025 at the latest, generation capacity that started commercial
production before 4 July 2019 and that emits more than 550 g of CO2 of fossil
fuel origin per kWh of electricity and more than 350 kg CO2 of fossil fuel
origin on average per year per installed kWe shall not be committed or receive
payments or commitments for future payments under a capacity mechanism.
The emission limit of 550 g CO2 of fossil fuel origin per kWh of electricity and
the limit of 350 kg CO2 of fossil fuel origin on average per year per installed
kWe referred to in points (a) and (b) of the first subparagraph shall be
calculated on the basis of the design efficiency of the generation unit meaning
the net efficiency at nominal capacity under the relevant standards provided
for by the International Organization for Standardization.
By 5 January 2020, ACER shall publish an opinion providing technical
guidance related to the calculation of the values referred in the first
subparagraph.’
Regulation (EU) 2019/943 includes Art. 22(5) says that ‘Member States that apply capacity mechanisms on 4 July
a grandfathering clause for capacity mechanism 2019 shall adapt their mechanisms to comply with Chapter 4 without prejudice
contracts concluded before 31 December 2019. to commitments or contracts concluded by 31 December 2019.’
Regulation (EU) 2019/943 requires that Art. 26(1) states that ‘capacity mechanisms other than strategic reserves
capacity mechanisms are open for cross-border and where technically feasible, strategic reserves shall be open to direct
participation. cross-border participation of capacity providers located in another Member
State, subject to the conditions laid down in this Article.’ Art. 26(3) adds that
‘Member States shall not prevent capacity which is located in their territory
from participating in capacity mechanisms of other Member States.’
 How to ensure adequate investment in power plants? 151

Section of this chapter, topic and relevant Relevant article

regulation
Regulation (EU) 2019/943 sets out the process Art. 26(7) states that‘for the purposes of providing a recommendation to
for the calculation of the maximum entry transmission system operators, regional coordination centres established
capacity for participation by foreign capacity in pursuant to Article 35 shall calculate on an annual basis the maximum entry
capacity mechanisms and the roles of the RCCs capacity available for the participation of foreign capacity. That calculation
and the TSOs. shall take into account the expected availability of interconnection and
the likely concurrence of system stress in the system where the mechanism
is applied and the system in which the foreign capacity is located. Such
a calculation shall be required for each bidding zone border.
Transmission system operators shall set the maximum entry capacity available
for the participation of foreign capacity based on the recommendation of the
regional coordination centre on an annual basis.’
Regulation (EU) 2019/943 sets the Art. 26(11) lists the different methodologies, common rules and terms to be
methodologies, rules and terms related developed by ENTSO-E and submitted to ACER. It states that ‘by 5 July 2020
to cross-border participation in capacity the ENTSO for Electricity shall submit to ACER:
mechanisms. (a) a methodology for calculating the maximum entry capacity for
cross-border participation as referred to in paragraph 7;
(b) a methodology for sharing the revenues referred to in paragraph 9;
(c) common rules for the carrying out of availability checks referred to in point
(b) of paragraph 10;
(d) common rules for determining when a non-availability payment is due;
(e) terms of the operation of the registry as referred to in point (a) of
paragraph 10;
(f) common rules for identifying capacity eligible to participate in the capacity
mechanism as referred to in point (a) of paragraph 10.’
PART III

How to put the citizen at the centre of the energy

transition?
8. How to put the citizen at the centre of the
energy transition?
Leonardo Meeus with Athir Nouicer

In this chapter we answer two questions. First, why did we start this new paradigm? Second,
how is this change of paradigm being implemented?

8.1 WHY DID WE START THIS NEW PARADIGM?

In this section, we discuss the extent to which the markets for electricity have delivered their
promise of better service and cheaper prices. We examine electricity bills, the energy transi-
tion and the political issues around the energy transition.1
First, electricity bills. The main components of electricity bills are the energy component,
the network component, and taxes and levies. The energy component is the wholesale market
price plus a retail margin. On average, EU wholesale prices have been steadily decreasing.
However, this is not always translated into a decrease in retail prices as not all countries have
well-functioning retail markets. The average EU household electricity bill increased by 28.2
per cent during the period 2008–2018. Other components of the electricity bill have been
increasing too. The network component increase reflects investments in networks necessary
for the integration of renewable energy resources. The taxes and levies, on top of value-added
tax, form the component that has increased the most. This has been in order to recover the
subsidies that have been provided to support the energy transition. These include support for
the large-scale deployment of renewable energy projects and energy efficiency measures.
They may also include the costs of social measures against energy poverty. On average in the
EU, the three main components of household electricity bills each make up about a third of the
total bill. In the same period, the average price for industrial customers only increased by 1.4
per cent. These customers have often contributed less or have been exempted from the taxes
and levies that drove up household bills. They have also had more opportunities to benefit
from the support mechanisms for renewable energy. Many industrial consumers invested in
the co-generation of heat and power, and in solar and wind parks. Some even own and operate
private distribution networks on their industrial sites. Until recently, households had fewer
opportunities to invest in the energy transition, and the opportunities were often limited to
privileged households with PV rooftops.
Second, the energy transition. The EU energy and climate targets for 2030 are more ambi-
tious than those for 2020. The greenhouse gas emission target increases from 20 per cent
to 40 per cent (reduction from the 1990 level), the renewable energy target increases from
20 per cent to 32 per cent (energy from renewable sources in the gross final consumption)
and the energy efficiency target increases from 20 per cent to 32.5 per cent (savings against
153
154 Evolution of electricity markets in Europe

a baseline energy consumption scenario established in 2007). The new European Commission
(2019–2024) has the ambition to achieve climate neutrality by legislating for a net zero green-
house gas emission target for 2050, and it might even bring forward this ambition to 2030. The
electricity sector is one of the sectors that has been partly decarbonized to achieve the 2020
targets. To achieve the 2030 and 2050 targets, this decarbonization will need to accelerate
so that other sectors can switch to electricity to decarbonize. The transport sector can then
be decarbonized with electric vehicles and other competing technologies, while the building
sector can be decarbonized with heat pumps as well as other competing technologies. A com-
bination of carbon pricing and subsidies will need to be used to push this forward, which will
put additional pressure on electricity bills.
Third, political issues. Increasing electricity bills for households, and bigger household bills
in general, have already created unrest in several countries. The unrest that has attracted the
most media attention is the Gilet Jaunes movement in France, which is voicing the frustration
of those that feel left behind. At the other end of the spectrum, there is a growing movement
of climate activists who are concerned that we are not going fast enough. School strikes and
climate marches initiated by young activists in several European countries have been all over
the media. The Clean Energy Package was developed and negotiated in the period up to 2019
and speaks to both forces in society. Citizens are invited to take ownership of the energy tran-
sition. Putting citizens at the centre of the energy transition is the new paradigm.

8.2 HOW IS THIS CHANGE OF PARADIGM BEING

IMPLEMENTED?

In this section, we explain how the new paradigm has been translated into an enabling regula-
tory framework. We first discuss individual and collective consumer rights, and then focus on
the actors that will play a key role in helping consumers exercise their rights and opportunities.

8.2.1 Individual and Collective Consumer Rights

Below, we discuss the concepts of self-consumption and collective self-consumption through

energy communities, and then the consumer’s entitlement to a smart meter with a dynamic
retail contract.2
First, self-consumption. It is true that in most countries consumers can already generate
and store electricity and easily manage their energy consumption. Nevertheless, there are still
some barriers in the current design of the retail market that may prevent consumers from fully
benefiting from such opportunities. Directive (EU) 2019/944 states that consumers should be
allowed to become active customers, or ‘prosumers’. They should be able to self-consume
what they produce to the extent that their production is aligned with their consumption, and
if it is not aligned they should be able to sell their production at wholesale prices or at prices
offered by retailers. This was already the case in most countries in Europe, but not all of them.
Directive (EU) 2019/944 calls for a phase-out of net metering. In countries with net metering,
consumers could deduct their production from their consumption even if the two were not
aligned. It meant that they were not paying for the network while they were still using it by
injecting energy in some periods and taking it back in others. Abolishing net metering is a very
first step towards more cost-reflective distribution network tariffs (see Chapter 4). No new
 How to put the citizen at the centre of the energy transition? 155

rights for schemes that do not account separately for the electricity fed into and consumed
from the grid are to be granted after 31 December 2023.
Second, collective self-consumption through energy communities. Not every household has
the necessary time, knowledge, financial means or space to become a prosumer. Energy com-
munities can help households access all or at least some of the prosumer benefits. Directive
(EU) 2018/2001, better known as the revised Renewable Energy Directive (RED II), requires
all countries to develop a regulatory framework for renewable energy communities (RECs).
These communities can be set up by consumers in the same apartment building, block, street,
neighbourhood or region to jointly invest in renewable energy projects. Countries can choose
how they define these communities in terms of proximity, but they should enable collective
self-consumption within a renewable energy community. Some form of energy coopera-
tive already existed in most countries, but a clear regulatory framework was often lacking.
Directive (EU) 2019/944 requires countries to develop a regulatory framework for citizen
energy communities (CECs). These communities can mobilize citizens in joint initiatives that
go beyond jointly investing in renewable energy projects. They can take over the role of other
market parties, such as retailers and aggregators. CECs can be local, like RECs, but they can
also be national. Many countries already have green retailers that are organized as coopera-
tives of citizens operating at the national level.
Third is consumer entitlement to a smart meter with a dynamic retail contract. The target of
reaching an 80 per cent roll-out of smart meters by 2020 will not be achieved in all the coun-
tries in Europe. Having a smart meter enables dynamic retail contracts, but such contracts have
not always been available in countries with smart meters. Directive (EU) 2019/944 continues
to encourage a smart meter roll-out, but it is not explicitly required. Countries that decide
not to go forward with a roll-out after conducting a cost–benefit assessment must in any case
ensure that consumers that want a smart meter can get one within a few months of requesting it
and at a reasonable price determined by the energy regulator. National regulatory frameworks
are to ensure that consumers with a smart meter are then entitled to a dynamic retail price
contract. Dynamic retail contracts have been defined as retail contracts with a temporal gran-
ularity at least equal to the market settlement period. Following Regulation (EU) 2019/943,
the settlement period is converging towards 15 minutes. Directive (EU) 2019/944 also defines
minimum functionalities for smart meters in accordance with Commission Recommendation
2012/148/EU. These include the provision of information on actual times of use, cyber secu-
rity and consumer privacy protection, consumption data availability on consumer request and
remote-control functionalities. At the time of writing, remote-control functionalities are only
available in five countries. By giving market parties or the system operator remote-control
over their meter or appliances, consumers can earn money for the flexibility they can provide
to the system.

8.2.2 Actors Expected to Enable the New Paradigm

In what follows, we show why distribution system operators (DSOs), transmission system
operators (TSOs), retailers and aggregators have a key role to play in enabling the new
paradigm.
First, DSO and TSOs. These will increasingly compete for the use of flexibility between
balancing and congestion management, which will require coordination. With the introduction
156 Evolution of electricity markets in Europe

of rooftop PVs, batteries, electric vehicles and heat pumps, flexibility is increasingly available
in distribution grids and can be aggregated in so-called virtual power plants of decentralized
energy resources. In Chapter 5, we showed that TSOs started to change balancing markets to
create a level playing field between the traditional and new sources of flexibility. Following
Directive (EU) 2019/944, countries also have to create an enabling regulatory framework
for DSOs to procure flexibility services from so-called flexibility service providers. DSOs
have to develop multi-annual investment plans that consider flexibility as an alternative to
network expansion. Some DSOs have already started to procure flexibility services in newly
established flexibility markets. Others have started to countertrade locally in intraday markets.
Note that this is very similar to redispatching actions by TSOs. Coordination is therefore
needed between TSO balancing actions, TSO congestion management actions in transmission
networks and DSO congestion management actions in distribution networks. How this will
work is an open issue.3
Second, retailers and aggregators. Retailers can become aggregators that help their cus-
tomers valorize their flexibility as balancing service providers for TSOs and/or flexibility
service providers for DSOs and TSOs. Retailers did not always do this because it conflicted
with their interests as commodity suppliers. Helping customers valorize flexibility can indeed
imply selling less volume. New players have therefore emerged that focus on the aggregation
business. They either had to become retailers themselves or ask existing retailers to access
their clients. Directive (EU) 2019/944 requires all countries to develop an enabling regulatory
framework for independent aggregators to operate next to retailers. This includes the possi-
bility of operating without consent from the retailer and an arrangement to organize compen-
sation between the retailer and the aggregator when one inflicts costs on the other. Similar to
the provisions for retailers, aggregators also have to provide data to their customers, enable
switching and provide clear terms and conditions in their contracts.
The future of retail markets is very much an open issue. There used to be one retailer
active at each connection point. Next to the traditional commodity retailer there will now be
independent aggregators interacting with the same customers. Retailers risk becoming backup
solutions for the supply of a commodity. Energy communities are alternative suppliers, and
peer-to-peer supply solutions are also emerging. Peer-to-peer platforms allow consumers to
source their energy from producers or prosumers of their choice. How the competition between
new and existing players in the retail market will be organized in the future is an open issue.

8.3 CONCLUSION

In this chapter about how to put the citizen at the centre of the energy transition, we have
answered two questions.
First, why did we start this new paradigm? Households electricity bills have increased
more than industrial ones. The energy transition is accelerating, with 2030 and 2050 targets
that are more ambitious than the 2020 targets were. Decarbonizing the electricity sector will
also help to decarbonize other sectors that could partly electrify, such as the transport sector
with electric vehicles and the building sector with heat pumps. Some citizens feel left behind,
while others want to accelerate the energy transition. The new paradigm is an answer to these
concerns. It increases opportunities for households to benefit from the energy transition.
 How to put the citizen at the centre of the energy transition? 157

Second, how is this change of paradigm being implemented? New individual and collec-
tive consumer rights through energy communities have been introduced. The actors that are
expected to enable the new paradigm are TSOs, DSOs, retailers and aggregators. Prosumers
and energy communities will increasingly participate in balancing markets and flexibility
markets to valorize their flexible resources. Aggregators can help them access these markets.
The future of retail is uncertain, with competition from energy communities and peer-to-peer
trading.

NOTES
1. ACER and CEER (2019a) provide a decomposition of household and industry electricity bills. The
EU average level for the different components is reported together with figures for country capitals
over the period 2008–2018. These figures are updated every year in the annual market monitoring
report.
2. CEER (2019) gives new and developing practices for collective self-consumption and energy
communities and analyses their regulatory implications. ACER and CEER (2019b) provide more
information on the status of smart meter roll-outs and the availability of dynamic retail contracts in
Europe.
3. In Nouicer and Meeus (2019) we list the different flexibility pilot projects and European initiatives
that are being implemented. In Schittekatte and Meeus (2020) we analyse four pioneering flexibility
market projects and show that they give a different answer to six key questions: Is the flexibility
market integrated in the existing sequence of EU electricity markets? Is the flexibility market
operator a third party? Are there reservation payments? Are the products standardized? Is there
TSO–DSO cooperation over the organization of the flexibility market? Is there DSO–DSO coop-
eration over the organization of the flexibility market? The market design options for TSO–DSO
coordination are being discussed at the EU level. CEDEC et al. (2019) provide different options for
coordination between system operators.

REFERENCES
ACER and CEER (2019a), ‘Market Monitoring Report 2018 – Electricity and Gas Retail Markets
Volume’.
ACER and CEER (2019b), ‘Market Monitoring Report 2018 – Consumer Empowerment Volume’.
CEDEC, ENTSO-E, GEODE, E.DSO and EURELECTRIC (2019), ‘TSO–DSO Report: An Integrated
Approach to Active System Management’.
CEER (2019), ‘Regulatory Aspects of Self-Consumption and Energy Communities’, Ref:
C18-CRM9_DS7-05-03.
Nouicer, A. and L. Meeus (2019), ‘The EU Clean Energy Package (Ed. 2019)’, FSR Technical Report.
Schittekatte, T. and L. Meeus (2020), ‘Flexibility markets: Q&A with project pioneers’, Utilities Policy,
63, 101017.
158 Evolution of electricity markets in Europe

8A.1 ANNEX: REGULATORY GUIDE

Table 8A.1 Regulatory guide

Section of this chapter, topic and Relevant articles

relevant regulation
Section 8.2.1
Directive (EU) 2019/944 defines the Art. 2(8) defines an active customer as ‘a final customer, or a group of jointly acting
concept of active customers and gives final customers, who consumes or stores electricity generated within its premises
final customers rights to act as active located within confined boundaries or, where permitted by a Member State, within
customers. other premises, or who sells self-generated electricity or participates in flexibility or
energy efficiency schemes, provided that those activities do not constitute its primary
commercial or professional activity.’
Art. 15(1) states that ‘Member States shall ensure that final customers are entitled to
act as active customers without being subject to disproportionate or discriminatory
technical requirements, administrative requirements, procedures and charges, and to
network charges that are not cost-reflective.’
Art. 15(2) lists the rights that shall be ensured to active customers. It states that
‘Member States shall ensure that active customers are:
(a) entitled to operate either directly or through aggregation;
(b) entitled to sell self-generated electricity, including through power purchase
agreements;
(c) entitled to participate in flexibility schemes and energy efficiency schemes;
(d) entitled to delegate to a third party the management of the installations required
for their activities, including installation, operation, data handling and maintenance,
without that third party being considered to be an active customer;
(e) subject to cost-reflective, transparent and non-discriminatory network charges that
account separately for the electricity fed into the grid and the electricity consumed
from the grid, in accordance with Article 59(9) of this Directive and Article 18 of
Regulation (EU) 2019/943, ensuring that they contribute in an adequate and balanced
way to the overall cost sharing of the system;
(f) financially responsible for the imbalances they cause in the electricity system; to
that extent they shall be balance responsible parties or shall delegate their balancing
responsibility in accordance with Article 5 of Regulation (EU) 2019/943.’
Directive (EU) 2019/944 requires Art. 15(4) states that ‘Member States that have existing schemes that do not account
a phasing out of net metering. separately for the electricity fed into the grid and the electricity consumed from the
grid, shall not grant new rights under such schemes after 31 December 2023. In any
event, customers subject to existing schemes shall have the possibility at any time to
opt for a new scheme that accounts separately for the electricity fed into the grid and
the electricity consumed from the grid as the basis for calculating network charges.’
 How to put the citizen at the centre of the energy transition? 159

Section of this chapter, topic and Relevant articles

relevant regulation
Directive (EU) 2018/2001 provides A REC is defined in Art. 2(16) as ‘… a legal entity:
a definition of a renewable energy (a) which, in accordance with the applicable national law, is based on open and
community (REC). voluntary participation, is autonomous, and is effectively controlled by shareholders
or members that are located in the proximity of the renewable energy projects that are
owned and developed by that legal entity;
(b) the shareholders or members of which are natural persons, SMEs or local
authorities, including municipalities;
(c) the primary purpose of which is to provide environmental, economic or social
community benefits for its shareholders or members or for the local areas where it
operates, rather than financial profits.’
Directive (EU) 2019/944 provides A CEC is defined in Art. 2(11) as‘… a legal entity that:
a definition of a citizen energy (a) is based on voluntary and open participation and is effectively controlled by
community (CEC). members or shareholders that are natural persons, local authorities, including
municipalities, or small enterprises;
(b) has for its primary purpose to provide environmental, economic or social
community benefits to its members or shareholders or to the local areas where it
operates rather than to generate financial profits; and
(c) may engage in generation, including from renewable sources, distribution, supply,
consumption, aggregation, energy storage, energy efficiency services or charging
services for electric vehicles or provide other energy services to its members or
shareholders.’
Directive (EU) 2019/944 provides A dynamic electricity price contract is defined in Art. 2(15) as ‘… an electricity supply
a definition of a dynamic electricity contract between a supplier and a final customer that reflects the price variation in
price contract. the spot markets, including in the day-ahead and intraday markets, at intervals at
least equal to the market settlement frequency.’
Directive (EU) 2019/944 ensures Art. 11(1) states that ‘Member States shall ensure that the national regulatory
that customers are offered a dynamic framework enables suppliers to offer dynamic electricity price contracts. Member
electricity price contract. States shall ensure that final customers who have a smart meter installed can request
to conclude a dynamic electricity price contract with at least one supplier and with
every supplier that has more than 200 000 final customers.’
Directive (EU) 2019/944 includes Art. 11(2) says that ‘Member States shall ensure that final customers are fully
a provision to inform consumers of the informed by the suppliers of the opportunities, costs and risks of such dynamic
costs and risks of dynamic electricity electricity price contracts, and shall ensure that suppliers are required to provide
contracts. information to the final customers accordingly, including with regard to the need to
have an adequate electricity meter installed. Regulatory authorities shall monitor
the market developments and assess the risks that the new products and services may
entail and deal with abusive practices.’
Following Regulation (EU) 2019/943, Art. 8(4) states that ‘by 1 January 2021, the imbalance settlement period shall be
the market settlement period is 15 minutes in all scheduling areas, unless regulatory authorities have granted
converging towards 15 minutes. a derogation or an exemption. Derogations may be granted only until 31 December
2024. From 1 January 2025, the imbalance settlement period shall not exceed 30
minutes where an exemption has been granted by all the regulatory authorities within
a synchronous area.’
160 Evolution of electricity markets in Europe

Section of this chapter, topic and Relevant articles

relevant regulation
Directive (EU) 2019/944 continues to Art. 19(1) states that ‘in order to promote energy efficiency and to empower final
encourage a smart meter roll-out, but it customers, Member States or, where a Member State has so provided, the regulatory
is not explicitly required. authority shall strongly recommend that electricity undertakings and other
market participants optimise the use of electricity, inter alia, by providing energy
management services, developing innovative pricing formulas, and introducing
smart metering systems that are interoperable, in particular with consumer energy
management systems and with smart grids, in accordance with the applicable Union
data protection rules.’
Art. 19(2) states that ‘Member States shall ensure the deployment in their territories
of smart metering systems that assist the active participation of customers in the
electricity market. Such deployment may be subject to a cost–benefit assessment which
shall be undertaken in accordance with the principles laid down in Annex II.’
Art. 19(5) adds that ‘where the deployment of smart metering systems has been
negatively assessed as a result of the cost–benefit assessment referred to in paragraph
2, Member States shall ensure that this assessment is revised at least every four years,
or more frequently, in response to significant changes in the underlying assumptions
and in response to technological and market developments. Member States shall notify
to the Commission the outcome of their updated cost–benefit assessment as it becomes
available.’
Directive (EU) 2019/944 requires Art. 19(3) states that ‘Member States that proceed with the deployment of smart
Member States proceeding with a smart metering systems shall adopt and publish the minimum functional and technical
metering system roll-out to adopt and requirements for the smart metering systems to be deployed in their territories,
publish the minimum functional and in accordance with Article 20 and Annex II. Member States shall ensure the
technical requirements of these systems, interoperability of those smart metering systems, as well as their ability to provide
in accordance with Art. 20 and Annex II output for consumer energy management systems. In that respect, Member States
of the same Directive. shall have due regard to the use of the relevant available standards, including those
enabling interoperability, to best practices and to the importance of the development
of smart grids and the development of the internal market for electricity.’
Art. 20 includes requirements on the functionalities of smart metering systems to
be followed in accordance with European standards and Annex II of Directive (EU)
2019/944. Among these requirements, smart meters ‘… shall accurately measure
actual electricity consumption and shall be capable of providing to final customers
information on actual time of use ...’
 How to put the citizen at the centre of the energy transition? 161

Section of this chapter, topic and Relevant articles

relevant regulation
In addition, validated historical consumption data will be made available to final
customers on request and at no additional cost. Also, non-validated near real-time
consumption data will be made available at no additional cost through a standard
interface or through remote access in a secure and easy way. The aim is to support,
inter alia, energy efficiency and demand response programmes. The smart metering
system security, data communication and privacy shall comply with Union security
rules. Smart meter operators will ensure that their devices account for the electricity
fed into the grid by active customers. Data related to the electricity injected will be
made available on request to the corresponding customers. They have the right to then
transmit their data to another party at no additional cost. Moreover, at the time of the
smart meter roll-out, final customers will receive appropriate advice and information
concerning their full potential, in accordance with Art. 20(f). Art. 20(g) adds that
‘smart metering systems shall enable final customers to be metered and settled at the
same time resolution as the imbalance settlement period in the national market.’
Commission Recommendation Section 42(g) states that every smart metering system shall ‘allow remote on/off
2012/148/EU sets common minimum control of the supply and/or flow or power limitation. This functionality relates to
functional requirements for smart both the demand side and the supply side. It provides additional protection for the
metering systems. consumer by allowing grading in the limitations. It speeds up processes such as when
moving home – the old supply can be disconnected and the new supply connected
quickly and simply. It is needed for handling technical grid emergencies. It may,
however, introduce additional security risks which need to be minimised.’
Directive (EU) 2019/944 entitles Art. 21(1) states that ‘where the deployment of smart metering systems has been
customers to request smart metering negatively assessed as a result of the cost–benefit assessment referred to in Article
systems in Member States where they 19(2) and where smart metering systems are not systematically deployed, Member
are negatively assessed. States shall ensure that every final customer is entitled on request, while bearing the
associated costs, to have installed or, where applicable, to have upgraded, under fair,
reasonable and cost-effective conditions, a smart meter that:
(a) is equipped, where technically feasible, with the functionalities referred to in
Article 20, or with a minimum set of functionalities to be defined and published by
Member States at national level in accordance with Annex II;
(b) is interoperable and able to deliver the desired connectivity of the metering
infrastructure with consumer energy management systems in near real-time.’
Section 8.2.2
Directive (EU) 2019/944 requires Art. 32(1)states that‘Member States shall provide the necessary regulatory framework
Member States to define the regulatory to allow and provide incentives to distribution system operators to procure flexibility
framework under which DSOs may services, including congestion management in their areas, in order to improve
acquire flexibility services. This will be efficiencies in the operation and development of the distribution system. In particular,
through a market-based process, unless the regulatory framework shall ensure that distribution system operators are able to
the regulatory authorities establish that it procure such services from providers of distributed generation, demand response or
is not economically efficient. energy storage and shall promote the uptake of energy efficiency measures, where
such services cost-effectively alleviate the need to upgrade or replace electricity
capacity and support the efficient and secure operation of the distribution system.
162 Evolution of electricity markets in Europe

Section of this chapter, topic and Relevant articles

relevant regulation
Distribution system operators shall procure such services in accordance with
transparent, non-discriminatory and market-based procedures unless the regulatory
authorities have established that the procurement of such services is not economically
efficient or that such procurement would lead to severe market distortions or to higher
congestion.’
Directive (EU) 2019/944 requires Art. 32(2) states that ‘Distribution system operators, subject to approval by the
DSOs to establish flexibility service regulatory authority, or the regulatory authority itself, shall, in a transparent and
specifications in a process including participatory process that includes all relevant system users and transmission system
all relevant stakeholders. The Directive operators, establish the specifications for the flexibility services procured and, where
invites DSOs and TSOs to coordinate appropriate, standardised market products for such services at least at national level.
over efficient, reliable and secure The specifications shall ensure the effective and non-discriminatory participation of
operation of their network. They should all market participants, including market participants offering energy from renewable
exchange data and information to ensure sources, market participants engaged in demand response, operators of energy
optimal use of the resources connected storage facilities and market participants engaged in aggregation ...’
to their networks.
RegardingDSO–TSO coordination, Art. 32(2) adds that ‘… Distribution system
operators shall exchange all necessary information and shall coordinate with
transmission system operators in order to ensure the optimal utilisation of resources,
to ensure the secure and efficient operation of the system and to facilitate market
development. Distribution system operators shall be adequately remunerated for
the procurement of such services to allow them to recover at least their reasonable
corresponding costs, including the necessary information and communication
technology expenses and infrastructure costs.’
Directive (EU) 2019/944 requires Art. 32(3) states that ‘the development of a distribution system shall be based on
DSOs to develop distribution network a transparent network development plan that the distribution system operator shall
development plans, with a possible publish at least every two years and shall submit to the regulatory authority. The
derogation for small DSOs. These network development plan shall provide transparency on the medium and long-term
plans are to include a trade-off between flexibility services needed, and shall set out the planned investments for the next
system expansion and the use of five-to-ten years, with particular emphasis on the main distribution infrastructure
flexibility services. DSOs will consult which is required in order to connect new generation capacity and new loads,
all relevant system users during the including recharging points for electric vehicles ...’ Network development plans are
development process. to include a trade-off between network development and alternative technologies. Art
32(3) adds that these plans ‘… shall also include the use of demand response, energy
efficiency, energy storage facilities or other resources that the distribution system
operator is to use as an alternative to system expansion.’
The development of these plans will involve different relevant stakeholders. Art. 32(4)
states that ‘the distribution system operator shall consult all relevant system users
and the relevant transmission system operators on the network development plan. The
distribution system operator shall publish the results of the consultation process along
with the network development plan, and submit the results of the consultation and the
network development plan to the regulatory authority. The regulatory authority may
request amendments to the plan.’
 How to put the citizen at the centre of the energy transition? 163

Section of this chapter, topic and Relevant articles

relevant regulation
Art. 32(5) allows a derogation from the development of distribution network
development plans, to be decided by Member States, for ‘integrated electricity
undertakings which serve less than 100 000 connected customers or which serve small
isolated systems.’
Directive (EU) 2019/944 establishes Art. 17(3) states, in the framework of demand response through aggregation, that
a regulatory framework clarifying ‘Member States shall ensure that their relevant regulatory framework contains at least
the roles and responsibilities of the following elements:
(independent) aggregators. It aims to (a) the right for each market participant engaged in aggregation, including
remove the barriers they may face when independent aggregators, to enter electricity markets without the consent of other
entering the market. market participants;
(b) non-discriminatory and transparent rules that clearly assign roles and
responsibilities to all electricity undertakings and customers;
(c) non-discriminatory and transparent rules and procedures for the exchange of
data between market participants engaged in aggregation and other electricity
undertakings that ensure easy access to data on equal and non-discriminatory terms
while fully protecting commercially sensitive information and customers’ personal
data;
(d) an obligation on market participants engaged in aggregation to be financially
responsible for the imbalances that they cause in the electricity system; to that
extent they shall be balance responsible parties or shall delegate their balancing
responsibility in accordance with Article 5 of Regulation (EU) 2019/943;
(e) provision for final customers who have a contract with independent aggregators
not to be subject to undue payments, penalties or other undue contractual restrictions
by their suppliers;
(f) a conflict resolution mechanism between market participants engaged in
aggregation and other market participants, including responsibility for imbalances.’
9. Conclusion
Leonardo Meeus

In the evolution of electricity markets in Europe, day-ahead markets have received most atten-
tion. It took a long time to integrate these markets, but the process is almost complete. What
was considered impossible at the start has been achieved. Intraday markets have been slower
in their development, but they are becoming more important with the transition to renewable
energy. In balancing markets, the definition of standard products and the creation of European
platforms to exchange these products across borders is a relatively new ambition, but much has
been achieved in a relatively short time. Of course, the devil is in the details, so implementa-
tion in the coming years will need to be closely monitored.
Looking back, most of the market integration process so far has required horizontal coordi-
nation between transmission system operators in different countries. Transmission constraints
were first managed by defining transmission rights and organizing separate markets for these
rights. The constraints were then integrated into wholesale markets and balancing markets
because this proved to be the best approach. The calculation of transmission constraints has
also been reorganized. The main open issues are the availability of transmission capacity
across timeframes and the definition of bidding zones. We currently have large bidding zones
with structural congestion inside them, and changing them is a sensitive topic.
Looking forward, vertical coordination between transmission and distribution system
operators is becoming more important, and this coordination process has only just started.
Distribution network constraints have always been avoided by (over-)investing in distribution
networks. This approach is becoming too expensive due to the integration of wind, solar,
electric vehicles and heat pumps in distribution grids. The trade-off between distribution grid
expansion and using flexibility to reduce the need for investment is one of the main challenges.
Flexibility can come from smart charging of electric vehicles and batteries, smart heating of
homes and the use of many other smart appliances.
To allow for coordination between all the resources at different voltage levels, this flexi-
bility deep inside the grid needs to be priced efficiently. Academics are already thinking of
wholesale and balancing markets that incorporate distribution network constraints to produce
distribution locational marginal prices. Practitioners in Europe currently do not consider this to
be feasible. Other solutions are favoured, such as flexibility markets, smart connection agree-
ments and distribution tariff reforms. We know from experience that these solutions can help,
but they have their limitations. It seems inevitable that we will go in the direction of DLMP.
The question is how long it will take.

164
Index
access charges 68–9, 70, 74 CBCA see cross-border cost allocation (CBCA)
ACER (The (European Union) Agency for the agreements
Cooperation of Energy Regulators) 7, 9, CCRs see capacity calculation regions (CCRs)
12, 13, 18–24, 30, 31, 34, 35, 43–4, 46–7, CEP see Clean Energy Package (CEP)
52, 55–6, 58, 63, 65–7, 70, 73, 75, 79, CGM see common grid model (CGM)
81–2, 84–5, 91, 92, 94, 97, 100, 105–7, citizens, placing at centre of energy transition
122, 127–8, 134, 140, 141, 146, 150–151, 153–4, 155–6, 157, 158–63
157 Clean Energy Package (CEP) 3, 4, 5, 9, 13, 18,
ACER Regulation (recast of) 3, 5, 18, 21–3 20–23, 31, 44, 52, 56, 57, 115, 123, 127–8,
active customers 80, 154–5, 158, 161 139, 142, 154
aggregation 123, 126, 130, 156, 157, 158, 163 clock incident 87–8
automatic frequency restoration process (aFRP) common grid model (CGM) 53–4, 62, 64, 65,
93, 101, 107 114, 127
automatic frequency restoration reserves (aFRR) congestion management 42, 95, 96, 155–6, 162
84–5, 87, 90, 93–4, 96, 100–102, 106–7, see also capacity allocation and congestion
109 management guideline (CACM GL)
congestion rent 40–41, 78
balance responsibility 84–8, 89–95, 96, 110 congestion revenue 69, 79
balance responsible parties (BRPs) 88–9, 91, 104, connection charges 68, 69, 74
141, 149, 158, 163 connection network codes (CNCs) 8, 116–20,
balancing capacity exchange 94–5, 108–10 121–3, 125–6, 127–34
balancing capacity tenders 90–91, 96, 105–7 connection requirements see system operation and
balancing energy exchange see exchange of connection requirements
balancing energy, European platforms for consumer rights 154–5
balancing energy markets 90–91, 96, 106 cost–benefit analysis (CBA) 72, 73, 81
balancing markets 33, 89, 90, 91–5, 96, 136, 141, cost-reflective network charges 70–71, 74, 80, 96,
149, 156, 157, 164 154, 158
balancing pilot projects 97 cost sharing 30
bidding zones 12, 40–41, 42–7, 50–51, 55–6, 57, Council of European Energy Regulators (CEER)
60, 61, 62, 63, 65–6, 67, 73–4, 88, 89, 103, 14, 55, 58, 63, 67, 91, 97, 157
146–7, 151, 164 cross-border cooperation 11–12
border trade constraints 48–50, 51–7, 60, 61, cross-border cost allocation (CBCA) agreements
62–7 72–5, 82
Brexit 12–13 cross-border intraday market project (XBID) 34,
BRPs see balance responsible parties (BRPs) 35
cruise ship incident 112–13, 123
capacity allocation and congestion management
guideline (CACM GL) 8, 20, 23, 29–31, data exchange 52–5, 57, 65
33–4, 42–7, 51–4, 56–7, 62–6, 102, 109, day-ahead timeframes 28–35
127 demand connection network code (DC NC) 8, 24,
capacity calculation regions (CCRs) 52, 53, 57, 116, 117–20, 121–2, 125, 126, 129–34
63, 66 Directive 96/92/EC see First Directive
capacity mechanisms 135–7, 138, 139–40, 141, Directive 2003/54/EC see Second Directive
142, 145–51 Directive 2009/72/EC see Third Directive

165
166 Evolution of electricity markets in Europe

Directive (EU) 2018/844 see Energy Performance European Network of Transmission System
in Buildings Directive Operators for Electricity (ENTSO-E) 5, 6,
Directive (EU) 2018/2001 see Renewable Energy 7, 9, 13, 18–22, 24, 30, 53, 54, 56, 64–5,
Directive (RED II) 67, 69, 71, 72, 75, 80–81, 85, 87, 92, 94,
Directive (EU) 2018/2002 see Energy Efficiency 96–7, 111, 114, 115, 118, 124, 127–9,
Directive 139–40, 141, 146, 151
Directive (EU) 2019/944 see Electricity Directive European platforms for exchange of balancing
(Recast of) energy 93–4, 96, 100, 102, 106, 107–8,
Directorate-General for Competition (DG COMP) 110, 164
55–6, 67 European Single Market 2, 13, 138
distribution locational marginal pricing (DLMP) European treaties 2, 13, 16
58, 164 exchange of balancing capacity 94–5, 108–10
distribution system operators (DSOs) 5, 7, 9, exchange of balancing energy, European
17–18, 21–2, 24, 53, 54–5, 65, 80, 95, platforms for 93–4, 96, 100, 102, 106,
110, 113, 116, 118, 126, 131, 137, 155–6, 107–8, 110, 164
162–3 explicit auctions 25, 26–7, 32, 33–4, 47, 69

electricity balancing guideline (EB GL) 8, 20, financial transmission rights (FTRs) 32–3, 46
23, 84–5, 87, 89, 90, 91–2, 93–4, 95, 96, First Directive 3–4, 16–17, 18, 25, 34, 42
100–110 First Energy Package 3–4, 5, 12, 17–18
electricity bills 153, 154, 156 flexibility markets 95, 96, 155–6, 157, 164
Electricity Directive (Recast of) 3–4, 5, 79, 90, flow-based market coupling (FBMC) 50–51, 52,
105, 128, 147, 154–6, 158–60, 161–3 55, 57, 60, 61, 62, 72
electricity market integration 2, 11–13, 16–17, 25, forward capacity allocation guideline (FCA GL)
27, 35, 38–9, 54–5, 92–5, 96, 164 8, 20, 23, 32–3, 34, 44–6, 52–3, 54, 57,
electricity price area differentials (EPADs) 33 63–5
Electricity Regulation (recast of) 3, 5, 18, 20–24, frequency containment 85–6, 92, 95, 100, 106,
44, 52, 56, 62–7, 70, 79–80, 88, 89, 90, 109
94–5, 101, 103–6, 109, 115, 122–3, 127–8, frequency containment reserves (FCR) 85–6, 87,
134, 139–40, 141, 146–51, 155, 158–9, 91, 94, 95, 101, 106, 108–9
163 frequency-controlled reserves 85, 86
electricity transmission system operation frequency restoration 86–7, 92, 113
guideline (SO GL) 8, 20, 23–4, 52, 53, see also automatic frequency restoration
54, 57, 62–5, 84–6, 87, 94, 95, 100–102, process (aFRP); automatic frequency
108–10, 114, 118, 123, 127, 129–30 restoration reserves (aFRR); manual
emergency and restoration network code (ER NC) frequency restoration reserves
8, 24, 62, 127 (mFRR)
energy communities 155, 156, 157, 159 frequency restoration control error (FRCE) 101–2
Energy Efficiency Directive 5 frequency restoration process (FRP) 85, 86–7, 96,
energy-only markets 135, 142, 145 101–2, 106–7
Energy Performance in Buildings Directive 5 frequency restoration reserves (FRR) 87, 90, 93,
energy storage 70, 79, 90, 105, 122–3, 134, 135, 94–5, 102, 105, 106, 109
138, 146, 147, 149, 159, 162, 163
ENTSO-E see European Network of G-component (generator component) 69, 72,
Transmission System Operators for 79–80
Electricity (ENTSO-E) generation adequacy 80, 135, 137, 141, 142
EU electricity network codes 7–10, 13, 19–21, 22, generation and load data provision methodology
23–4, 31, 34–5, 56, 64, 122–3, 134 (GLDPM) 53, 54, 63–4, 65
EU legislative energy packages 4–7, 13, 16–24 grid user, different types of 116–17, 125–6
EU resource adequacy assessment 135, 139–40,
142, 146–8 harmonization of network tariffs 69–70, 74
EUPHEMIA (Pan-European Hybrid Electricity high voltage direct current (HVDC) 74, 125–6
Market Integration Algorithm) 29, 30–31 high voltage direct current network code (HVDC
NC) 8, 24, 116, 123, 125
 Index 167

historical privileges for trade across borders 25–6, missing money problem 91, 96, 135–6, 138, 142
34
N–1 redundancy principle 49, 57
IGMs see individual grid models (IGMs) national regulatory authorities (NRAs) 5, 7, 8–9,
imbalance netting 92–3, 95, 96, 107, 109 17–18, 20, 22, 25, 31–2, 34, 43, 46–7, 52,
imbalance settlement 88–9, 91, 96, 103, 141, 149 53, 56, 63, 65, 71–3, 82, 100, 107–8, 114,
imbalance settlement period (ISP) 44, 89, 96, 122, 125, 127, 134
104–5, 106, 141, 149, 159, 161 net transfer capacity (NTC) 50–51, 58, 62
implicit auctions 27, 32, 34, 41, 47 network codes see connection network codes
see also market coupling (CNCs); EU electricity network codes
independent system operators (ISOs) 48, 142–3 network tariffs 68, 69–70, 71–5, 78, 79–82
individual grid models (IGMs) 53–4, 64 nodal pricing 56–7, 58, 88
inter-TSO compensation (ITC) scheme 71–2, 74, Nominated Electricity Market Operators
75, 80 (NEMOs) 8, 20–21, 29–31, 42–4, 47
intraday auctions 33–4, 35, 47 Nordic Cooperation of Electricity Utilities
intraday capacity calculations 51, 54, 57, 62–3, (NORDEL) 84, 85, 86
127 Norway–Sweden case 73–4
intraday common grid models 64
intraday cross-zonal capacity 47 offshore power park module (OPPM) 116, 125,
intraday cross-zonal gate closure 46–7, 95, 102, 129, 133
109–10 Operational Planning Data Environment (OPDE)
intraday cross-zonal pricing 34, 47 54, 64
intraday markets 30, 33–4, 35, 38, 39, 42–4, 87, over the counter (OTC) trading 31, 38
95, 110, 156, 159, 164
intraday technical price limit 141, 149 peer-to-peer trading 156, 157
intraday trading services 29, 42 Physical Communication Network (PCN) 54
intraday transmission rights 33–4 physical transmission rights (PTRs) 32–3, 46
Italian blackout 49–50 power exchanges 27, 28–33, 35–6, 38–9, 40–41,
Italy–Greece case 74 42–3
ITC mechanism see inter-TSO compensation power-generating modules (PGMs) 116–22, 125,
(ITC) scheme 128–33
power park module (PPM) 116, 118, 120, 125,
Joint Allocation Office (JAO) 32, 44 129, 132
power plants investment see capacity mechanisms
KORRR methodology 54–5, 65 power system 10, 11, 13, 53, 64, 84, 87, 111, 118,
122, 123
load frequency control (LFC) 86, 87, 94, 101–2, priority lists 9, 19, 22, 25–6
108–10, 130 pro-rata rationing 25–6
load-shedding plans and controversies 136–7 proactive balancing 87, 96
loop flows 61 projects of common interest (PCIs) 72–3, 74–5,
80–82
manual frequency restoration reserves (mFRR)
84–5, 87, 90, 93–4, 96, 100, 102, 106–7, rationing of supply 136–7
109 RCCs see regional coordination centres (RCCs)
market-based allocation of transmission rights reactive balancing 87, 96, 102
26–7, 34 real-time energy pricing 89, 105
market coupling 27, 28–34, 35, 40–41, 42–3, 78 regional coordination centres (RCCs) 52, 53–4,
see also flow-based market coupling 57, 62, 63, 64, 66, 67, 94–5, 109, 114, 115,
(FBMC) 127, 128
market coupling operator (MCO) 25, 29–30, 34, regional security coordination initiatives (RSCIs)
40–41, 42–3, 60 51–2, 62, 113–14, 123
market splitting 25–6, 28, 29, 33 regional security coordinators (RSCs) 51–2, 53,
missing market problem 136, 142 57, 62–3, 64, 114–15, 124, 127, 128
regional system operation 113–15, 123
 168 Evolution of electricity markets in Europe

Regulation (EC) 713/2009 3, 18, 20–21 single intraday coupling (SIDC) 34, 42–4, 47
Regulation (EC) 714/2009 3, 18–24, 72, 79–80, solar eclipse incident 111–12, 115, 123
116 solidarity mechanism 10–11, 85–6, 95–6, 101
Regulation (EC) 1228/2003 3–4, 26, 34, 42, 69, synchronous power-generating module (SPGM)
72, 79–80 116, 118, 120, 125, 129, 132
Regulation (EU) 2018/1999 see Regulation on the system disturbance 111–12
Governance of the Energy Union system inertia 10, 11, 13, 118, 122
Regulation (EU) 2019/941 see Regulation on system operation and connection requirements
Risk-Preparedness 111–13, 115–20, 121–3, 125–6, 127–34
Regulation (EU) 2019/942 see ACER Regulation system operation guideline (SO GL) 8, 20, 23–4,
(recast of) 52, 53, 54, 57, 62–5, 85, 86, 87, 94, 95,
Regulation (EU) 2019/943 see Electricity 100–102, 108–10, 114, 118, 123, 127,
Regulation (Recast of) 129–30
Regulation (EU) No 347/2013 see
Trans-European Energy Network (TEN-E) tandem bicycle analogy 10–11
Regulation technical standards 115–20, 121, 122–3, 125–6
Regulation on Risk-Preparedness 3, 5, 115, 123, Tempus case 138
128 Ten-Year Network Development Plan (TYNDP)
Regulation on the Governance of the Energy 72, 75, 80
Union 5, 147 Terms and Conditions or Methodologies (TCM)
regulatory guides 16–24, 42–7, 62–7, 79–82, 8–9
100–110, 127–34, 145–51, 158–63 Third Directive 3–4, 17–18, 104, 125–6, 129
Renewable Energy Directive (RED II) 5, 155, 159 Third Energy Package 3–4, 5, 7, 8–9, 17–23, 129
requirements for grid connection of generators trade across borders 25–6, 28–31, 33–4, 35, 38–9,
network code (RfG NC) 8, 24, 116, 40–41, 42–7
117–18, 120, 121–2, 125, 126, 128–34 see also border trade constraints;
reserve products, terminology for 84–5 transmission rights
reserve replacement 85, 87, 100, 106, 108–9 Trans-European Energy Network (TEN-E)
reserves 11–12, 85–6, 87, 89–92, 93, 94–5, 96, Regulation 72–3, 74, 80–82
101, 102, 105, 106, 108–10 transit charges 71–2, 74, 80
see also automatic frequency restoration transit flows 61, 71–2
reserves (aFRR); frequency transmission lines 48–9, 55, 57, 72, 80, 95, 109
containment reserves (FCR); manual transmission planning 48–9, 57
frequency restoration reserves transmission rights 25–7, 32, 33–5, 42, 44–6, 50,
(mFRR) 69, 94, 78, 136, 164
resource adequacy see EU resource adequacy transmission system operators (TSOs) 5–7, 8, 9,
assessment 11, 13, 17–18, 20–21, 25, 26, 28, 30, 32–3,
risk preparedness 115 34–5, 38–9, 42–7, 48–52, 53–5, 56, 57,
see also Regulation on Risk-Preparedness 60, 62–7, 68–9, 71–2, 73, 74, 78, 79–80,
RSCs see regional security coordinators (RSCs) 82, 84, 86–8, 89, 90, 92–5, 96, 100–110,
111–13, 114–15, 118–20, 121–2, 123, 126,
scarcity pricing 91–2 127–34, 137, 138, 141, 145, 151, 155–6,
Second Directive 3–4, 17, 18 157, 162
Second Energy Package 3–4, 5, 7, 12, 17–18, 26 TSOs see transmission system operators (TSOs)
security of supply 2, 16, 105, 136–8, 141, 147
sharing balance responsibility between system Union for Coordination of Transmission of
operators 84–8, 95–6 Energy (UCTE) 49–50, 84, 85, 86, 112
sharing network investment costs between
countries 72–5 virtual border calculations see border trade
sharing of reserves 11–12, 94–5, 108–10 constraints
single day-ahead coupling (SDAC) 28–9, 31, 34, volume coupling 28–9, 35
43–4
single energy market 12, 16
Single European Act 1986 2, 16 zonal congestion pricing approach 50

View publication stats